Gateway Fundamentals
Gateway Fundamentals
Gateway Fundamentals
1: Gateway Fundamentals
Students Training Guide
S150-2708-00
October 2007
Copyright Notice Copyright 2007 IBM Corporation, including this documentation and all software. All rights reserved. May only be used pursuant to a Tivoli Systems Software License Agreement, an IBM Software License Agreement, or Addendum for Tivoli Products to IBM Customer or License Agreement. No part of this publication may be reproduced, transmitted, transcribed, stored in a retrieval system, or translated into any computer language, in any form or by any means, electronic, mechanical, magnetic, optical, chemical, manual, or otherwise, without prior written permission of IBM Corporation. IBM Corporation grants you limited permission to make hardcopy or other reproductions of any machine-readable documentation for your own use, provided that each such reproduction shall carry the IBM Corporation copyright notice. No other rights under copyright are granted without prior written permission of IBM Corporation. The document is not intended for production and is furnished as is without warranty of any kind. All warranties on this document are hereby disclaimed, including the warranties of merchantability and fitness for a particular purpose. Note to U.S. Government UsersDocumentation related to restricted rightsUse, duplication or disclosure is subject to restrictions set forth in GSA ADP Schedule Contract with IBM Corporation. Trademarks The following are trademarks of IBM Corporation or Tivoli Systems Inc.: IBM, Tivoli, AIX, Cross-Site, NetView, OS/2, Planet Tivoli, RS/6000, Tivoli Certified, Tivoli Enterprise, Tivoli Ready, TME. In Denmark, Tivoli is a trademark licensed from Kjbenhavns Sommer - Tivoli A/S. Microsoft, Windows, Windows NT, and the Windows logo are trademarks of Microsoft Corporation in the United States, other countries, or both. UNIX is a registered trademark of The Open Group in the United States and other countries. C-bus is a trademark of Corollary, Inc. in the United States, other countries, or both. Java and all Java-based trademarks are trademarks of Sun Microsystems, Inc. in the United States, other countries, or both. Lotus is a registered trademark of Lotus Development Corporation. PC Direct is a trademark of Ziff Communications Company in the United States, other countries, or both and is used by IBM Corporation under license. ActionMedia, LANDesk, MMX, Pentium, and ProShare are trademarks of Intel Corporation in the United States, other countries, or both. SET and the SET Logo are trademarks owned by SET Secure Electronic Transaction LLC. For further information, see http://www.setco.org/aboutmark.html. Other company, product, and service names may be trademarks or service marks of others. Notices References in this publication to Tivoli Systems or IBM products, programs, or services do not imply that they will be available in all countries in which Tivoli Systems or IBM operates. Any reference to these products, programs, or services is not intended to imply that only Tivoli Systems or IBM products, programs, or services can be used. Subject to valid intellectual property or other legally protectable right of Tivoli Systems or IBM, any functionally equivalent product, program, or service can be used instead of the referenced product, program, or service. The evaluation and verification of operation in conjunction with other products, except those expressly designated by Tivoli Systems or IBM, are the responsibility of the user. Tivoli Systems or IBM may have patents or pending patent applications covering subject matter in this document. The furnishing of this document does not give you any license to these patents. You can send license inquiries, in writing, to the IBM Director of Licensing, IBM Corporation, North Castle Drive, Armonk, New York 10504-1785, U.S.A. Printed in Ireland.
Table of Contents
Unit 1: Gateways Overview
Course Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3 A Telecommunication Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-4 Possible Network Management Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-5 Gateway Defined . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6 Example: Raw Data File Extract from Nokia OMC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-8 Parser Intermediate Format File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-9 Loader Input Format File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-10 Gateway Framework Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-12 Main Components and Use of Files and Directories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-13 Framework Connection to Vendor-Specific Engines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-14 Gateway Framework Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-15 Other General Features of the Gateway Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-18 Vendor Gateways . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-20 Gateway Directories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-21 Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-24 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-25
Table of Contents
StatisticsConfig.pm: File Statistics Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . StatisticsConfig.pm: File Statistics Entries (l) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . StatisticsConfig.pm: File Statistics Entries (ll) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . StatisticsConfig.pm: File Statistics Entries (lll) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Statistics Configuration: Block Stats Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . StatisticsConfig.pm: Block Statistics Entries (ll) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . StatisticsConfig.pm: Counter Statistics Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . StatisticsConfig.pm: Counter Statistics Entries (l) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . StatisticsConfig.pm: Counters Statistics Entries (ll) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Notifications: Alarm Messages Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alarm Notifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Notifications: NPR Event Commands Message Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . Event Notifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . NotificationsConfig.pm: Standard Entries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . NotificationsConfig.pm: Example Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Parallel Processing: Configuration Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Parallel Processing: Example Configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Post Parser Rules Supporting Parallel Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3-34 3-36 3-37 3-39 3-41 3-44 3-46 3-47 3-49 3-51 3-52 3-53 3-54 3-54 3-56 3-58 3-59 3-61 3-63 3-65 3-66
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Table of Contents
MERGE_RECORDS: Mandatory Entries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MERGE_RECORDS: Optional Entries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MERGE_RECORDS: Example Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PERLIZE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PERLIZE: Mandatory Entry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PERLIZE: Optional Entries (l) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PERLIZE: Optional Entries (ll) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PERLIZE: Example Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PIF_2_OUTPUT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PIF_2_OUTPUT: Optional Entries (l) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PIF_2_OUTPUT: Optional Entries (ll) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PIF_2_OUTPUT: Example Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PIF_REMOVE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PIF_REMOVE: Example Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SPLIT_RECORDS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SPLIT_RECORDS: Mandatory Entries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SLIT_RECORDS: Optional Entries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SPLIT_RECORDS: Example Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UNPEGGER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UNPEGGER: Time and Duration Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UNPEGGER: Mandatory Entries (l) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UNPEGGER: Mandatory Entries (ll) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UNPEGGER: Mandatory Entries (lll) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UNPEGGER: Mandatory Entries (lV) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UNPEGGER: Optional Entries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UNPEGGER: Sample Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Student Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4-39 4-40 4-41 4-42 4-44 4-45 4-47 4-48 4-49 4-50 4-51 4-52 4-53 4-54 4-55 4-56 4-57 4-58 4-59 4-62 4-64 4-66 4-68 4-70 4-71 4-72 4-73 4-74
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
III
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Table of Contents
IV
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
1-1
Course Objectives
IBM Software Group | Tivoli software
Course Objectives
Upon completion of this course, you will be able to :
Understand what a Gateway is and its usefulness. Describe the most important features of the Gateway Framework. Learn how to easily install a Gateway Framework and the necessary vendor Gateways. Customize Gateways using the most important Pre-Parser and Post-Parser rules.
Introduction
During this course you will learn the basic concepts related to Gateways, a key component
for Performance Manager for Wireless, Metrica Performance Manager, and Network Performance Reporting .
You will also be introduced to technical aspects related to the Gateway Framework, ranging from Gateways installation to customization using standard components. This course has been designed to cover the most important aspects of the Gateway Framework when related to Performance Manager for Wireless (PMW); however, the Framework is quite versatile and is also used with Network Performance Reporting (NPR) and Metrica Performance Manager (MPM), and has the possibility to be used with Prospect as well. Therefore, throughout this course you will also find a number of references for these other products as well.
1-2
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Objectives
IBM Software Group | Tivoli software
Objectives
Upon completion of this unit, you will be able to:
Understand what a Gateway is and its usefulness. Understand and describe a Loader Input Format (LIF) file. Describe the main features and constituent blocks of the Gateway Framework. Understand how the Gateway Framework and the vendor Gateways interact.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
1-3
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
A Telecommunication Network
IBM Software Group | Tivoli software
ISDN/PSTN Network
C VLR
BTS
Abis
(G)MSC
PCU
Gb Gn
Gi Gi
Internet
GGSN
Gp
X.25
SGSN
CGF
Other PMLN
To understand what is a Gateway, you must first understand why a Gateway is necessary. When you consider the task of understanding how a Telecommunications Network functions, the first thought that comes to mind is how complex the network can be. The slide shows an example of a block diagram depicting a GPRS network. The complexity arises when the equipment of each vendor uses different software versions, which generate performance counters and data file formats that are not standardized. This makes the task of analyzing the performance of even a single piece of equipment very difficult - and the task of analyzing the network as a whole almost impossible.
1-4
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
CS Domain
Public Voice Network Are there IBM Software Group | Lotus software database
HLR MSC VLR
xRAN
PS Domain
SGSN
GGSN
problems?
Internet
If you consider the task of analyzing the whole network, the effort is compounded by dealing with multiple vendors and equipment types. As illustrated, in any telecommunication network, the performance analysis questions are quite varied in their technical and business aspects, as well as on the equipment they affect. The matter becomes complicated because the counters are not standardized. Sometimes even the telecommunication company may struggle with defining performance measurements. Finally the information needs to be compiled and presented in a suitable manner to network operations managers and operators.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
1-5
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Gateway Defined
IBM Software Group | Tivoli software
Gateway Defined
Counters Format 1
Gateways Framework
Vendor GW 1
Counters Format 2
Vendor GW 2
To resolve these issues, the Network Performance Management products rely on software called Gateway. A Gateway is basically written in Perl software that translates the counters made available by the vendor equipment into various possible ouput files which can be used by the Network Performance Management software. The most common output format is called Loader Input Format (LIF). These output files are used as input information to the Network Performance Management software, which will ultimately be responsible for generating the reports and graphs that are useful to the operators. Considering the fact that one Gateway type per counter data format type is needed, these can be any combination of file structures such as XML, CSV, ASCII, Binary, Inverse Binary, and so forth, and file internal configuration (which depends on vendor equipment type and software), you would need to have a large number of gateways created to cover ALL these combinations. This would be a quite demanding task, both in time and resources, if you try to create one Gateway for each counter data format type that is created in the market. To save time and resources, the Gateway Framework was created.
The Gateway Framework provides a mechanism to create:
1-6
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
- standard counter processing components that are common to any gateway regardless of the counter data format type; - PERL script extensions to handle vendor specific gateways.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
1-7
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
. . .
03MAR00,C01AM,1000,1100,BZRDOR3,8331,1811,ZRDOR,6,12,M01BM,82,0,2,1,0,0,0,82, 16,30,46,0,0,1,0,2,0,0,0,0,0,0,0,30,0,0,0,0,0,0,2,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,30,0,0,0,0,0,0,0,16,0,0,0,0,0,0,0,0,0,2,0,0,0,0,0,0,0,0,0,0,0,0,0; 03MAR00,C01AM,1000,1100,BZRDOR4,8332,1811,ZRDOR,6,12,M01BM,65,0,3,2,1,0,0,65, 24,40,64,0,0,1,0,0,0,0,0,0,0,0,0,40,0,0,0,0,0,0,3,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,40,0,0,0,0,0,0,0,24,0,0,0,0,0,0,0,0,0,3,0,0,0,0,0,0,0,0,0,0,0,0,0; 03MAR00,C01AM,1000,1100,BJILUK1,8835,1811,JILUK,7,4,M01BM,31,0,0,0,0,0,0,31,2,22,24,0,0,0,0,0,0,0,0,0,0,0,0,2 2,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,22,0,0,0,0,0,0,0,2,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0; END,p_nbsc_traffic;
The slide shows a portion of a raw data file from a Nokia Operation Maintenance Center (OMC). The raw data file was extracted from a network component.
1-8
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
The slide shows an example of what the file looks like internally to the Gateway Framework and vendor Gateways. This is a Parser Intermediate Format (PIF) file, which is a temporary format for the performance data, used internally by the Gateway to transfer data between different stages and rules. You will learn about this in more detail when the Gateway Framework architecture is discussed.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
1-9
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
#%npr Cell Channel Report #%% Parser file sw1aday1-jan-96 { day 01jan96 start 0915 end 0930 bsc sw1a cell_tch { cellid 01001 tch 8 traff 2.34 bids 34 } cell_tch { cellid 01002 tch 8 traff 3.45 bids 48 } cell_sdcch { cellid 01001 sdcch 8 traff 0.87 bids 45 } cell_sdcch { cellid 01002 sdcch 8 traff 0.87 bids 67 } }
10
Loader Input Format (LIF) is the standard file format created by the Parser scripts for input to the Network Management Performance software. It is designed to be easy to produce and to reflect some of the structuring encountered in data obtained from any network component. The Loader Input Format is a structure of nested blocks. This provides a simple method for mirroring the likely structure of raw performance data files, which is also well defined and easily parseable. An input format file contains a standard header file. This provides the Loader with the assurance that the file is valid and not some extraneous file that has ended up in the Loader transfer directory. The Loader rejects any file that does not start with the characters #%npr, followed by a title (if it is an PMW/NPR input), or #%metrica (if it is a MPM input). The title is optional, and is used to help identify the data in log files. The input file consists of a sequence of report blocks, enclosed in brackets. Each block contains either identifier or value pairs that associate a value with a field name, or more blocks. Nested blocks inherit all identifier or value associations defined in the outer blocks. The closing of the brackets around each block triggers the Loader to process and to load data into the database. Some blocks, usually the innermost blocks, are named, that is, they have block names. A block name references a particular file, record type, or report in the performance data,
1-10
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
and is the starting point for mapping the data to the Network Management Performance software data model. The input format file lists the cell Traffic Channel (TCH) and Standalone Dedicated Control Channel (SDCCH) data for a specific Base Station Controller (BSC). The BSC, day and time period of the data are identified in the header and the file name is passed on to the log file of the Loader. The outer block defines the start and end collection times and the node identifier of the BSC. The next level of block nesting names the cells. Within the BSC block, blocks are named either cell_tch or cell_sdcch, and contain the cell identifier and the relevant traffic data for that cell. The layout of the file is irrelevant, providing white space separates the lexical tokens, although indentation of nested blocks makes it easier to read. The identifier for the network element must be in the innermost block. These identifiers do not have to be unique across your entire network, but you must be able to define a unique identifier for each active element in the Loader input map.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
1-11
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
11
The purpose of the Gateway Framework is to provide a standard framework for the transfer, parsing, manipulation, and presentation of performance data from remote network elements to the IBM Network Performance Management applications. The raw performance data is typically transferred from the operators of Operations and Maintenance Centers (OMC). This data is exported at fixed periods, for example, 15 minutes. The format of this data is typically bulky and in a non-standard format. The framework consists of a set of processing stages that perform different functions to output the required performance data. It includes standard tools for logging and recovery, as well as configurable rules for the transformation of data. The final output is then loaded to the IBM Network Performance Management application to provide user reporting and monitoring services.
1-12
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
GATEWAY FRAMEWORK
raw files
Transfer (IN)
Parser Post Transfer IBM Software Group | Lotus software(OUT) Engine Parser
LIF
raw files
raw files
PIF files
PIF files
LIF files
LIF files
IN_DIR
INT_DIR
OUT_DIR
12
The slide gives a brief overview of the Gateway Framework architecture showing the main components, and the files and directories used. The Gateway Framework consists of three main components: Transfer Engine (IN and OUT), Parser Engine and Post Parser Engine. These components interact mainly with various directories for reading, writing, and storing of files. The Transfer component is split into separate IN and OUT phases, which are run before and after parsing of the data.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
1-13
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Basic Framework
Framework Engines
13
If you look at the framework in a three dimensional way, you can see that the basic framework is actually connected to the vendor-specific engines through the configuration files of the Gateway Framework. By configuring Gateway Framework files, you can add various vendor-specific modules to the basic framework. You will learn in this course how this can be done. The following pages give a brief description of each component of the Gateway Framework.
1-14
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Transfer (IN)
Parser Engine
14
Transfer (IN): This optional stage allows the configuration of the transfer of raw performance data from remote servers, typically the OMC. It supports the scp, ftp, rcp, and local cp file transfer mechanisms. Multiple instances can be configured, so files can be transferred from multiple destinations. Parser Engine: The engine stage parses the raw performance data files, which are either vendor or third party standard formats, producing the data in PIF format. This processing is performed by a vendor-specific engine rule, which allows the Gateway Framework to be used for multiple vendors, using different engine rules to match the vendor data. The engine stage writes the PIF data to an intermediate directory, or to a common memory store. For example, to parse Nokia ASCII format data, the NOKIA_ASCII engine rule is used, which creates PIF files.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
1-15
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Post-Parser
Transfer (OUT)
15
Post-Parser: The Post-parser stage processes this data through multiple standard and vendor-specific Post-Parser rules, to transform data for efficient loading in accordance with the Performance Management (PM) systems load map. Examples of common functions performed within the Post-Parser stage are: Joining of PIF files. Insertion of hierarchy data from a secondary data source, for example, insertion of Global System for Mobile Communications (GSM) hierarchy data into performance data based on Base Station Transceiver System (BTS) and CELL keys. Accumulation of counter values across rows, for example, accumulation of cell counters across a BSC.
The Post-Parser rules can produce two types of output: PIF files, which are used as input to the next Post-Parser rule on the processing chain. Output files, which are saved to a separate directory. These are the final, transformed performance files for loading into the PM system. These can be one of the following file types:
1-16
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
XML files
A vendor Gateway is usually released with a Post-Parser configuration that meets the standard requirements for mapping of data to the PM system. Ownership of the complete customer configuration and solution, which usually consists of further data manipulation rules, is the responsibility of deployment professionals. Transfer (OUT): Once the output files have been created, they can be transferred to a remote server for loading, as required. The Transfer engine once again handles this data transfer. It is configured for the transfer of files from the local server to a remote server. The configuration of this stage is optional, because this stage may be unnecessary, or handled by other external tools.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
1-17
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
16
Logging and Auditing The Gateway Framework provides standard audit and logging tools for Vendor Gateways.
Crash Recovery The Gateway Framework manages the recovery from a crash due to not valid input file, so that the Gateway will not continually fail on a restart.
Backlog Processing The Gateway Framework can drip feed files in the oldest first order through the Gateway, allowing controlled recovery from a backlog situation.
Parallel Processing Engine and Post-Parser rules can be configured to process groups of files in parallel.
1-18
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Memory Caching of Intermediate Data To further improve performance, intermediate data being passed through various stages of transformation may be cached in memory, rather than on disk. Note that this now only applies to the post-parsing rules. The engine portion of the Gateway will always generate PIF files into the intermediate directory. This facilitates multiple processes being applied to the engine stage, while retaining the benefits of PIF caching at the Post-Parser stage.
Saving of Parsed, Intermediate, and Load Data Files from various stages of processing can be configured to be saved to directories for traceability.
Compression of Input and Stored Data Compression input data can be handled automatically, and stored data files can be compressed to reduce storage requirements.
Automated Data Transfer The Framework contains support for the transfer of data both to and from the local server using a number of standard transfer protocols.
Notification The Gateway Framework provides a Notification module for dispatching alarm and statistical event activity messages to the Network Performance Manager application notification system.
Statistics Gathers file, block and counter statistics on the data processed by the vendor gateway.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
1-19
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Vendor Gateways
IBM Software Group | Tivoli software
Vendor Gateways
OMC GGSN SGSN RNC BSC MSC Nokia Siemens Motorola
Alcatel-Lucent
. . .
17
The Gateway Framework provides a standardized architecture for the collection, parsing and presentation of vendor-specific data formats. The vendor Gateway consists of a number of Engine and Post-Parser modules to perform the manipulation of vendor-specific data. As mentioned previously, it is integrated into the framework through the configuration files. You can have a large variety of vendor Gateways, but they tend to be quite reusable thanks to the configurable architecture of framework.
1-20
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Gateway Directories
IBM Software Group | Tivoli software
Gateway Directories
vstart parsersrc IN_DIR INT_DIR
IBM OUT_DIR
Storage Directories
18
This slide lists the main Gateway directories. The configuration and use of these directories will be further described in Unit 3: Gateway Framework Configuration and Management, however, the following list provides a quick description of them: vstart: This directory contains the startup script, gateway_start.sh, as well as the configuration files for the Gateway and vendor parsing functions. This may be a specific version number for example, r9.1, if a particular configuration applies to a vendor data version. parsersrc: This directory contains the vendor-specific engine and Post-Parser rules. IN_DIR (raw directory): This directory contains the raw performance data files to be parsed by the Gateway. It can contain a directory hierarchy, if required, and the Gateway will navigate through the directory hierarchy. These files may be transferred using either the transfer stage in the Gateway Framework or an external script. INT_DIR (PIF directory): This directory is used for the output of PIF files while the Gateway is running. The output PIF files of one rule are passed to the next rule through this directory. If required, PIF files can also be stored here in-between Gateway iterations, when they must be saved for certain rules.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
1-21
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
OUT_DIR: This directory is used for the output of the final performance data files from the Gateway. These files can be loaded into the PM system either directly from these directories or transferred using the transfer engine to a remote server. Storage directories: These directories are used to optionally store the raw, intermediate, and output files. Typically a different directory is configured for the storage of each file type.
1-22
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Review Questions
IBM Software Group | Tivoli software
Review Questions
Describe the objectives of a Gateway Framework. What is the difference between the Gateway Framework and a vendor Gateway? What is a LIF file and how is it structured? How many vendor Gateways can be offered to a client?
Describe the five features from the Gateway Framework you believe are the most important to a client.
19
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
1-23
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Student Exercise
IBM Software Group | Tivoli software
Student Exercise
Access the server where your Gateway Framework is installed and determine:
The vendor Gateways available The main directories used by the Gateway Framework
Write down the Software GroupArchitecture, substituting IBM Gateway Framework | Lotus software the names in the comprising blocks (Post-Parser, Transfer (IN), and so forth) by the actual directories path where these engines are located.
20
1-24
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Summary
IBM Software Group | Tivoli software
Summary
You should now be able to:
Understand what a Gateway is and its usefulness. Understand and describe a LIF file. Describe the main features and constituent blocks of the Gateway Framework. Understand how the Gateway Framework and the vendor Gateways interact.
21
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
1-25
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
1-26
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
2-1
Introduction
In this unit you will learn the steps required to install any vendor Gateway, as well as their underlying Framework and Perl scripts. You will be also able to perform a quick installation in class, with help from your instructor.
Objectives
IBM Software Group | Tivoli software
Objectives
Upon completion of this unit, you will be able to:
Install Perl Install the Gateway Framework Install vendor Gateways
23
2-2
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
IBM Software Group | Lotus software 4) Run Configure and follow or set the parameter when prompted.. $ ./Configure
5) Run make. $ make 6) Login as root. Run make install. $ make install 7) Test the compiled Perl; It should print Hello world! on the screen. $ <path of the build directory>/bin/perl e print Hello world!;
24
Unlike previous releases, version 3.x of the Gateway Framework is now decoupled from the vendor Gateway. The Gateway Framework and the vendor Gateway are now supplied and installed separately. This has a number of advantages for release management and version support. It also reduces duplication for installation of multiple Gateways on a single server. Before proceeding with the vendor Gateway installation, the following tasks must be completed: Install a valid version of Perl. Install the Gateway Framework.
Gateway installations should only use the appropriate version of Perl, downloaded and installed from the Perl website. The product release notes and install guides contain current information about how to install and test the installation. To do that, you should perform the steps described in the picture above. Notice that CPAN modules are required to be previously installed on the machine, otherwise the system shall not work. An important point should be raised on step 4: this build shall be set as a non-threading Perl. Please specify the installation directory for Perl to be built otherwise the Perl binaries will be install in the default location and may overwrite existing Perl binaries which may have already been installed in the system.
Copyright IBM Corp. 2007 IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
2-3
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
25
Before installing and configuring a vendor Gateway, the Gateway Framework must be downloaded and installed. The Gateway Framework consists of the set of common packages and modules used by all vendor Gateways. Therefore, it only needs to be installed once on a server and referenced by multiple vendors. This package is included in the product media package as part of the installation process.
The installation procedure for the Gateway Framework is documented in the applicable product installation guide and release notes.
2-4
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
26
After the steps in the previous slide have been completed, the applicable Vendor Gateway can be downloaded and installed. To install the Vendor Gateway: Download the tar file from the Gateways web site to the installation directory. This should be a new directory (even if there is an existing one available) in order to: Prevent previous versions being overwritten Provide a backup in case there are problems with an upgrade
For this example, for an alcatel-nss installation, the following directory will be used: /metrica/gw/alcatel-nss DO NOT FORGET to create this directory manually and perform the operations from inside it, since this is NOT automatically created. Gunzip the release Untar the release
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
2-5
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
The installation is now complete - it is recommended that you have the applicable product installation guides and Vendor Gateway release notes as reference during your installation process to check if there are any specific details. After the initial system configuration is performed the Gateway will run.
2-6
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Student Exercise
IBM Software Group | Tivoli software
Student Exercise
Follow the guidance of your instructor and install the Gateway Framework and the available vendor Gateway.
27
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
2-7
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Summary
IBM Software Group | Tivoli software
Summary
You should now be able to:
Install Perl Install the Gateway Framework Install vendor Gateways
28
2-8
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
3-1
Introduction
This unit describes the configuration of the Gateway, its various stages, and the system management through the provided scripts.
Objectives
IBM Software Group | Tivoli software
Objectives
Upon completion of this unit, you will be able to:
Perform the basic Gateway Framework configuration. Start the Gateways using cron. Set up transfer configuration. Set up parser engine configuration. Set up Post-Parser configuration. Set up statistics configuration. Configure Gateway notifications. Configure Gateways for parallel processing.
30
3-2
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
31
The general configuration of the Gateway is performed in two files: The first one is the gateway_start.sh. Set the location of the basic installation parameters such as Perl and the Gateway Framework installation. The gateway_start.sh script starts the vendor Gateway. It contains a number of configuration entries that can be set either within the file, or by using environment variables. The configuration entries are: PERL5: After the correct Perl installation has been downloaded, the path to the Perl 5.6.1 installation should be set according to the product installation guide. For Performance Management for Wireless, it is suggested as:
/appl/virtuo/gways/perl
GATEWAY_FRAMEWORK: The path to the Gateway Framework installation should be set according to the product installation guide. For Performance Management for Wireless, it is suggested as:
/appl/virtuo/gways
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-3
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
LOG_FILE: This variable is used to detail the name and location to write the log file to. Is set by default to: ./log
LOG_LEVEL: The log level of the Gateway is set to 5, the highest level by default. This is useful during the installation and configuration of the Gateway, and the level should be reduced to 3 for the final commissioning. You may also see below the details for each log level: 5: More Detailed Information 4: Warning 3: Data Error [from level 3 - including - downwards, parser would stop processing] 2: Configuration Error 1 1: Configuration Error critical 6: Used main for debugging by the developers, might not be removed in actual source code.
RELEASE: The name of the release directory containing the vendor specific configuration data, is set to vstart by default. This can be set to different values if different configuration versions are being supported. PROPERTY_FILE: The name and location of the properties file. By default this points to the standard properties file contained in vstart. Besides those, it is also important to know that there are two variables for HP-UX that point to the location of UNIX shared libraries required by the Gateway Framework : LD_LIBRARY_PATH, and SHLIB_PATH. These need to be correctly set in these systems before any Gateway operation, and are important only for products that can be actually installed in a HP-UX machine - Network Performance Reporting for example.
3-4
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
32
The second file responsible for general Gateway configuration is the Properties file: Properties: Use to set general, system level configuration, such as the directories for the raw, intermediate, and output data. The property entries can include environment variables in the form of ${var}. The properties file configuration entries are in the following format: <name>=<value> where <name> is the name of the property, and <value> the value assigned to it. The property file entries are: AUDIT_FILE: The name of the file to write the audit trail messages to. By default it is set to: AUDIT_FILE=audit
DISK_FREE: The Gateway can be configured to check for free minimum disk space on all used directories on startup. If this space is not available the Gateway will log a warning with the required and available free space and exit. The free space required can be specified as a percentage: DISK_FREE=10% or number of MB free DISK_FREE=300 MB
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-5
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
This setting can be used to prevent the Gateway failing during processing as it exhausts disk space while writing intermediate, or output files. Set the value to 0 to ignore. IN_DIR: The root directory for the raw performance data files. The VendorEngine rules will read this directory for files to process. Subdirectories can be created to separate different types of files, if appropriate. INT_DIR: The intermediate directory for PIF files. If PIF data is being cached to memory, then only PIF files being kept between Gateway iterations will be visible in this directory (non-debug mode).
3-6
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
33
OUT_DIR: The output directory for the final performance output files created by the Gateway. INPUT_STORAGE_DIR: The directory to store the raw files after processing. This is optional, set it to 0 if it is not required. INTERMEDIATE_STORAGE_DIR: The directory to store the PIF files created during the Gateway run. This is optional, set it to 0 if it is not required. Even if in memory PIF caching is being used, they will be output as files to this directory, depending on the configuration. OUTPUT_STORAGE_DIR: The directory to store the performance LIF files. DEBUG: Switch on or off debugging. With debugging on, the raw and PIF files will not be removed from their directories to allow the performance of the Gateway to be traced. DEBUG=debug Switches on debugging DEBUG=0 Switches off debugging.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-7
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
COMPRESS: Compress data in the storage directories. The default is 0, no compression. The entry: COMPRESS=0 Turns off compression and: COMPRESS=true Turns on compression
COMPRESS_TOOL: The compression utility to use for compressing files. This should be set to the path where compress or gzip is installed. REMOTE_COMPRESS_TOOL: This entry is similar to the COMPRESS_TOOL, but it is meant for the remote server. The compression utility specified here must match with the one specified in COMPRESS_TOOL. This will enable the cpio channel to compress the data before transferring.
3-8
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
34
PIF_MODULE: The PIF storage option to use for the storage of PIFs during processing. There are the following two options: PIF_MODULE=PIF_Handler This is the standard mode, where PIF data is written and read from files in the intermediate directory. PIF_MODULE=PIF_Cache
The PIF data is written and read to local memory rather than disk. PIF files will still be written to disk, if required between Gateway iterations, and to the INTERMEDIATE_STORAGE_DIR, if configured. In memory PIF caching has performance advantages over disk based PIF storage. Note that PIF caching no longer applies to the engine stage of processing and only starts at the Post-Parser stage. MAX_PIF_LOCAL_MEMORY: The maximum local memory that should be used by the PIF_Cache. If it exceeds this value, then a warning will be printed in the log file. In general, the memory based PIFs require about 2.3 times the amount of space of the disk based PIFs. Initially in this file, all that must be set to get the Gateway up and running is the data directories. Other values can be configured as required. The properties value settings can also include environment variables in the form ${var}.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-9
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
35
Typically the Gateway is run at specified intervals, for example, every 15 minutes. This is controlled by using the standard unix tool, cron. The Gateway has the functionality to ensure that a second Gateway will not start to process the same set of raw performance files. This can occur in a backlog situation, where the amount of data to process takes longer than can be handled in the 15-minute period. Above you can see an example of a crontab starting two gateways, on intervals different from the standard 15 minutes, to show that this is also feasible and that the 15-minute period is not mandatory.
3-10
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
IN_DIR
36
The Transfer stage controls the transfer of files, both onto and off the local server. Raw files are transferred into the local input directory for processing, and performance LIF files may be transferred from the local server to a remote destination for loading. This stage is optional and may not be necessary, or may be performed by external tools. In this case the default TransferConfig.pm, which is empty (that means, contains no actual rules), can be used. Multiple transfer rules can be configured within this file. Raw files can be copied from multiple remote destinations, and similarly the performance data can be transferred to multiple remote destinations. The mandatory, optional, and protocol specific configuration entries will be described in the following slides. See the following example TransferConfig.pm file.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-11
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
37
3-12
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
38
RULE_DESC: A description that will be logged to the audit file. This is useful for tracking the rule performance. DIRECTION: Set to IN or OUT. Specifies whether files are being transferred to or from a remote destination. Raw files for processing would be specified with: DIRECTION => IN and performance files for transfer to a remote server DIRECTION => OUT
PROTOCOL: The Protocol to use for transfer. This can be set to ftp, scp, or rcp. If performing a transfer to or from the local server, set the protocol to rcp, and the TransferEngine will automatically detect that it is a local directory and use the appropriate local commands. HOST: The hostname or IP address of the remote destination LOCAL_DIR: The local directory to copy files to when DIRECTION is IN, or the local directory to copy files from when DIRECTION is OUT REMOTE_DIR: The remote directory to copy files from when DIRECTION is OUT, or the remote directory to copy to when DIRECTION is IN. TIMESTAMP_FILE: The file used to save the timestamp of last file copied in or out.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-13
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
NOTE: This file must be unique for each rule entry. See further notes on the timestamp file. DEPTH: The directory depth to follow down when copying files from a remote server. If DEPTH is set to 0, then only files in the base remote directory will be copied, if DEPTH is set to 1, then subdirectories that are one down also be copied. Note this directory structure will be replicated in the local directory. INPUT_FILE_DESCRIPTION: A hash containing the regular expressions used to match different file types for copying. Only files matching the entries in this hash will be copied.
3-14
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
39
STATISTICS_FILENAME: The name of the file to output the statistics to. This file has the date and time the transfer stage ran appended, and will contain statistics on the total time, number of files copied, and size of files for each entry in INPUT_FILE_DESCRIPTION. OUTPUT_FORMAT: The format to use to write the transfer statistics. Can be any of the supported output formats, LIF_Writer, CSV_Writer, XML_Writer, or Prospect_Writer. PRODUCE_PIF: If set to true, statistics will be output in PIF format. They can then be joined to file statistics, if required. ENABLE_PING: Whether or not to attempt to ping the remote host before initiating the transfer session. PING_PROTOCOL: The protocol to use for the ping request. This can be set to icmp, udp, or tcp depending on the ping type required. Note for icmp pings the Gateway must be running as root user. PING_RETRY_ATTEMPTS: The number of times to reattempt the ping if it fails. The default is 0. RETRY_INTERVAL: The number of seconds to wait between ping attempts. The default is 5 seconds.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-15
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
ENABLE_LOCAL_COMPRESSION: Whether or not to compress the local files copied in or out. Files transferred in will be compressed once they have been transferred. Files transferred out will be compressed before transfer to reduce the network load. NUMBER_OF_FILES_TO_PROCESS: If required, the files can be drip fed into the Gateway. If set to a value, the TransferEngine will not copy in more than this number of files on each iteration. This only applies to files being copied onto the local server.
3-16
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
40
DELETE_ORIGINAL: Whether or not to delete the original file on the remote server, in the case of DIRECTION IN, or on the local server, in the case of DIRECTION OUT. Defaults to 0, do not delete. TMP_FILENAME_EXTENSION: The temporary extension to use on files while they are being copied. This prevents incomplete files being read by a process during transfer. Defaults to .pt. TIMEOUT: Numerical value in seconds for file transfer timeout. Defaults to 20 seconds if not set. OVERWRITE_FILES: Whether or not to overwrite the existing file on the remote server, in the case of DIRECTION OUT, or on the local server, in the case of DIRECTION IN. Defaults to 0, do not overwrite. Set to TRUE to overwrite.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-17
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
ENABLE_REMOTE_COMPRESSION RETRY_ATTEMPTS
41
For the File Transfer Protocol (FTP): USER: The username to login to the remote server PASS: The password to login to the remote server TRANSFER_MODE: The Transfer mode can either be configured to ASCII or BINARY. The Transfer mode is set to binary by default.
For the SCP and RCP commands: PROTOCOL_PATH: The location of the protocol installation. This should be the path to where the ssh or rsh installation is, usually /usr/local/bin. This is used to run the ssh or scp, and the rsh or rcp commands used during the transfer process. BULK_TRANSFER: When this entry is set to TRUE or 1, the cpio command will be used to transfer files. The cpio command will create a private channel between the local host and the remote host before transferring files, thus increasing the performance when there is a large quantity of files to transfer. To further increase the performance, a new entry in the properties file for the remote server compression utility is required. All destination files will be overwritten and the drip feed mechanism will be ignored for the bulk transfer. Do not set BULK_TRANSFER, or set to 0 to disable the bulk transfer.
3-18
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
NOTE: The total size of files to be transferred is limited to 2 GB for bulk transfer. This is a limitation of the cpio tool imposed by XPG/4 and POSIX 2 standard. NOTE: ENABLE_LOCAL_COMPRESSION and ENABLE_REMOTE_COMPRESSION will work exclusively, where ENABLE_REMOTE_COMPRESSION will take precedence when DIRECTION IN, and ENABLE_LOCAL_COMPRESSION will take precedence when DIRECTION OUT. RETRY_ATTEMPTS: The number of times to reattempt the transfer if it fails. The default is 0. ENABLE_REMOTE_COMPRESSION: Whether or not to compress the remote files copied in or out before the are transferred.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-19
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
SCP and RCP: Creates the timestamp file in the remote directory on the remote server
42
This slide explains the usage of the timestamp file within the different protocols. As stated before the timestamp file should be unique for each rule entry. FTP: When configured with the ftp, the TransferEngine will create the timestamp file in the local directory on the local server. Within this file it stores the modification time of the last file, and the list of the last files copied. This is to ensure that if using drip feed of files, it will know which files with a certain modification time have already been copied. SCP and RCP: When configured with the scp and rcp commands, the TransferEngine will create the timestamp file in the remote directory on the remote server. The engine uses this file in combination with the find command to get a list of files which are newer than the timestamp file on the remote server. It then filters the files based on the INPUT_FILE_DESCRIPTION. The timestamp file does not contain any information, only the modification time of the file is used. Drip feed cannot be used with scp or rcp unless DELETE_ORIGINAL is also set to TRUE.
3-20
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Student Exercise
IBM Software Group | Tivoli software
Student Exercise
Check the following sample for an output and an input transfer configuration from TransferConfig.pm, and explain the expected transfer behavior of your Framework.
In your opinion, what can be done in this example to enhance the transfer capabilities of your Framework?Lotus software IBM Software Group |
43
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-21
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
LOCAL_DIR => "../data_dirs/input_d", REMOTE_DIR => "/spool/nokia/dumped_files", TIMESTAMP_FILE => "./lock_dir/.itimestamp", INPUT_FILE_DESCRIPTION => { BSC => 'bsc_.*\.raw', BTS => 'bts_.*\.raw', }, DELETE_ORIGINAL => 0, ENABLE_LOCAL_COMPRESSION => 'TRUE', OVERWRITE_FILES => 'TRUE', },
44
3-22
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
45
The configuration for the engine stage, where the raw performance data is parsed in the standardized PIF format, is contained in EngineConfig.pm. This configuration can consist of a number of rule entries, depending on the complexity of the vendor data and the number of formats supported. The configuration entries and values within each of these depends on the Vendor-Engine rule. You will see in the following slides the Mandatory and Optional configuration entries in this file.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-23
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
INPUT_FILE_DESCRIPTION
46
RULE_TYPE: The name of the vendor Engine that is being run. This must match the name of the Perl module in the parsersrc directory.
RULE_TYPE => NOKIA_ASCII
INPUT_FILE_DESCRIPTION: A scalar or array specifying the regular expressions which the files must match. This is to ensure only the appropriate raw data files are passed to the Vendor-Engine rule for processing. Can be configured as either: INPUT_FILE_DESCRIPTION => BSC.*.raw or an array of different filenames INPUT_FILE_DESCRIPTION => [BSC.*\.raw, NSS.*\.raw]
3-24
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
RULE_DESC: A text string that will be logged to the audit file. This can be useful for tracing performance of the same engine rule.
RULE_DESC => Parse Nokia ASCII BSC raw performance data
NUMBER_OF_FILES_TO_PROCESS: The number of files to be processed through the Engine and post parser in a single run. This is used to drip feed files through the system in a backlog scenario. ORDER_OF_FILES: The order in which files are to be processed. These options are: YOUNGEST_FIRST OLDEST_FIRST DIRECTORY_ORDER FILENAME_ASCENDING FILENAME_DESCENDING
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-25
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
FILENAME_HEADER_FIELDS: A set of new counters extracted from the filename, which are passed to the vendor engine rule. In the following example the date and time are being extracted from the filename.
FILENAME_HEADER_FIELDS => { DATE => BSC\.(\d{8}).* TIME => BSC\.\d{8}(\d{2}:\d{2}) }
DIRECTORY_HEADER_FIELDS: Similar to FILENAME_HEADER_FIELDS, with fields extracted from the directory path to the raw file. INPUT_DIR_DEPTH: The maximum depth in the raw input directory to follow. For example:
INPUT_DIR_DEPTH => 1
The engine will process all files in the raw directory, and files located one directory level further down. DO_NOT_DELETE: Do not delete the raw files or move them to the storage directory.
3-26
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
48
After the Engine stage is complete, the intermediate PIF files are passed to the Post-Parser stage. The Post-Parser stage is configured in UserConfig.pm. The rules are processed in the order in which they appear in UserConfig.pm, so the output of one Post-Parser rule can feed into the next. Two types of Post-Parser rules can be configured: Standard: These are the Post-Parser rules shipped as standard with the Gateway Framework. These rules fulfill most of the standard functions required to create the output performance data. The next unit contains a full description of these standard rules. Vendor Specific: Where the standard rules do not meet the user requirements for output, vendor specific Post-Parser rules are designed to meet the specific needs of the vendor data. These are shipped with a vendor Gateway.
The specific configuration entries for each Post-Parser rule depend upon the functionality it offers. However, all Post-Parser rules support some standard entries that are configured for each rule. In the next slides you will have a quick description of these standard entries.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-27
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
PRODUCE_PIF OUTPUT_FORMAT
49
RULE_TYPE: The Post-Parser rule to run. This must be the same as the name of the Perl module that implements the rule.
RULE_TYPE => JOIN
RULE_DESC: A text string that will be logged to the audit file. This can be useful for tracing operation of the same engine rule.
RULE_DESC => Join P_NBSC_TRAFFIC, P_NBSC_RES_AVAIL files,
3-28
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
OUTPUT_FORMAT: The Output Format module to use to write out the final data. For LIF data specify:
OUTPUT_FORMAT => LIF_Writer
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-29
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
OUTPUT_RECORD_KEY_DELIMITER
50
HEADER_COUNTERS_ORDER: The order in which to output the header counters. This entry can either contain a list of counter names, a hash of block names with a list of counter names for each block, or the filename to an ASCII file containing a list of counter names in a single column. For LIF_Writer, XML_Writer, and CSV_Writer, this only lists the counters that are required to be sorted and which appear at the start of the output block. All other counters in the output data will be sorted lexically. If the entry is not set, no sorting is performed. For Prospect_Writer, it is mandatory to provide this entry with a list of all the counters that are required to be sorted and appear in the output block. All other counters will be ignored. If the entry is not set, no counters will be output.
HEADER_COUNTERS_ORDER => [ qw(BSC DATE STARTTIME) ],
DATA_COUNTERS_ORDER: Similar to HEADER_COUNTERS_ORDER, the order in which to output the data counters. OUTPUT_RECORD_KEY: A list of data counters to form the record key for the first column of the output block. OUTPUT_RECORD_KEY_DELIMITER: A string to separate the counter names for the output record key. If not set, the default -#- is used.
3-30
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Student Exercise
IBM Software Group | Tivoli software
Student Exercise
Check the UserConfig.pm snapshot in the slide, it has been introduced before, and explain the expected Post-Parser behavior of your Framework.
In your opinion, what can be done in this example to enhance the Post-Parser capabilities of your Framework? IBM Software Group | Lotus software
51
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-31
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
52
The Gateway Framework supports four different types of statistics: File Statistics: Statistics on the raw performance data files processed and the output LIF files produced. Block Statistics: Statistics on the names, number of instances of each block name in the output, and the number of counters in each block. Counter Statistics: Statistics on the names of counters present in both the raw and LIF data. Transfer Statistics: Statistics on the raw performance files being transferred in, and the LIF files transferred from the server. See Transfer Configuration: TransferConfig.pm for more information.
Statistics serve two main purposes with the Gateway Framework: Verify the integrity and completeness of data being presented to the Gateway for parsing and the output data for loading to the PM system. Self reporting of the type, amount (number of files), and size (of files) being parsed by the Gateway.
All statistics are produced in a configurable output format so that they can be loaded directly into the PM system.
3-32
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Note that the file and block statistics form the basis of the event notification message being sent to the notification engine for dispatching. There is an overhead in running statistics on the raw and output performance data. To turn off statistics completely, remove or comment out the entries from the statistics configuration file, StatisticsConfig.pm. The entries relating to the different areas for statistics collection are described in the next slides.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-33
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
53
File statistics are collected for each iteration of the Gateway. These statistics are then summarized at a configurable interval, the summary period. For example, the Gateway may be running every 15 minutes, but the file statistics are summarized once a day. File Statistics collect the following statistics on the raw files parsed by the Gateway: Summary start date and time Summary end time Measurement type Number of raw files parsed (successfully, failed, 0 size) Total size of raw files (in KB) The number of files in each measurement period. This is a sliding scale, with the number of files for each 30-minute period for the first 2 hours, then hourly periods. The calculations are based on the time difference between the current time of parsing and the end time of the performance measurement file period. These values can therefore be used to detect backlog buildup.
3-34
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
For the output file data the following information is captured: The number of output files created The size of these output files
These statistics are stored in intermediate PIF files during the Gateway run and then, depending on the configured summary period, joined to create the final output file.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-35
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
| Lotus software
STATISTICS_HEADER_ENTRIES
54
STATISTICS_FILE_DESCRIPTION: The regular expressions listing the input raw files that will be matched by this rule. Statistics will be gathered for the set of files and grouped by type.
STATISTICS_FILE_DESCRIPTION => [HLR,VLR],
STATISTICS_HEADER_ENTRIES: A list of counter names and values to insert into the output statistics file.
STATISTICS_HEADER_ENTRIES => { GatewayID => 3GPP_XML, Network => NSS, Version => 1.2, },
3-36
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Lotus software
LIF_TYPE_MAPPINGS
55
STATISTICS_HEADER_MAPPINGS: This is a list of counter names to be mapped into the statistics output header. The hash keys are the header counter names to use in the output statistics header data. The hash values are the actual header counter names from the engine PIF output for one raw file processing whose values are mapped to the output statistics header data. Note that the Engine module is responsible for populating these and it is critical that this mapping works in order to collect file statistics. The entries StartDate, StartTime, EndTime, and Type are all mandatory and must be mapped.
STATISTICS_HEADER_MAPPINGS StartDate => StartDate, StartTime => StartTime, EndTime => EndTime, Type => Network_FileType }, => {
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-37
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
HEADER_INFO_FOR_STATS_FILE: This is a list of header values to use when creating the name of the statistics file. These can be from either the STATISTICS_HEADER_MAPPINGS or STATISTICS_HEADER_ENTRIES list.
HEADER_INFO_FOR_STATS_FILE => [ qw(GatewayID Type ParsingDate ParsingTime)]
STATISTICS_OUTPUT_DIRECTORY: This is the output directory to which to write the File Statistics files for collection. The directory must exist and and must have read and write privileges.
STATISTICS_OUTPUT_DIRECTORY => ../file_statistics/,
LIF_TYPE_MAPPINGS: This provides a map between the raw input files and the resulting output files, the LIF. Each entry consists of the type (derived in the STATISTICS_HEADER_MAPPINGS) being mapped to a series of Regular Expressions (REs) which match the LIF files that are produced by those raw file types (HLR and VLR, and so forth).
LIF_TYPE_MAPPINGS => { NSS => ['HLR','MSC'], BSS => 'BSC', },
3-38
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
56
STATISTICS_SUMMARY_PERIOD: The period (in minutes) to summarize file statistics. This allows file statistics to be accumulated independently of the scheduled interval of the Gateway.
STATISTICS_SUMMARY_PERIOD => 1440, # One day.
The gateway checks to see if the summary period has passed. If true, then the existing file statistics PIF files generated during the summary period are summarized and output using Post-Parser JOIN and ACCUMULATE rules. The file statistics current in the Gateway statistics buffers for the current run will not be included in the summary. These will be output in PIF format and summarized in the next summary. OUTPUT_FILENAME_START: An optional string to append to the output filename.
OUTPUT_FILENAME_START => File_Statistics,
DATA_KEY_NAME: A dummy key to insert into the PIF files which can then be used by the JOIN and ACCUMULATE rules when summarizing the results over the STATISTICS_SUMMARY_PERIOD.
DATA_KEY_NAME => FS_Key,
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-39
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
A full example of the rule configuration is contained in StatisticsConfig.pm, supplied with the vendor Gateway. Note that the file statistics included in the event notification message are based on the results for the just completed run of the Gateway, and do not rely on the summarized data.
3-40
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
57
The block statistics are collected on the output formatted files. The block statistics are output directly to the configured output format file. Note that the functionality has been updated to only output one block statistics file per run of the Gateway. Each block statistics output file contains: The start date and time The end time The measurement type
And a block in the block statistics output file for each block in the output data containing: The block name The number of blocks of this name The number of counters in this block
The example slide on this page shows a typical file that has Block Stats in it. In the next slides you will see the Block Statistics Configuration entries.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-41
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
58
STATISTICS_FILE_DESCRIPTION: This is a list of REs, one of which the LIF files must match, for statistics collection from that LIF. The following entry matches all output LIF files.
STATISTICS_FILE_DESCRIPTION => [.*\.lif],
STATISTIC_FILE_ENTRIES: This is the list of entries that will be extracted from the LIF filename, which can be subsequently used in the block statistics filename, or a header.
STATISTICS_FILE_ENTRIES => SUB_TYPE => '^(\w{3})', }, {
3-42
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
STATISTICS_HEADER_ENTRIES: This is a list of counter name or value attributes that will be printed to the LIF header.
STATISTICS_HEADER_ENTRIES GatewayID => 3GPP_XML, Network => NSS, Version => 1.2, MeasurementType => HLR, }, => {
STATISTICS_HEADER_MAPPINGS: These are the names of header entries from the output file to map to the statistics header. These are hash entries where the key represents the name of the counter that will be output in the block statistics output file and the value represents the name of the counter in the output file from which you are collecting statistics.
STATISTICS_HEADER_MAPPINGS => { StartDate => DATE, Time => TIME, },
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-43
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
STATISTICS_UNKNOWN_SUB_TYPE: This is a list of entries that have been extracted from the LIF filename, using STATISTICS_FILE_ENTRIES. These entries will then be used as part of the block statistics filename.
STATISTICS_UNKNOWN_SUB_TYPE => NULL,
STATISTICS_FILE_ENTRIES_FOR_HEADER: This is a list of entries that have been extracted from the LIF filename, using STATISTICS_FILE_ENTRIES. These entries will then be used as part of the block statistics filename.
STATISTICS_FILE_ENTRIES_FOR_HEADER => [qw(SUB_TYPE)],
HEADER_INFO_FOR_STATS_FILE: The values from the STATISTICS_HEADER_MAPPINGS and STATISTICS_HEADER_ENTRIES are used to create the filename. These can be used to ensure uniqueness of the block statistics filenames.
HEADER_INFO_FOR_STATS_FILE => [ qw(GatewayID MeasurementType DATE TIME)],
STATISTICS_BLOCKS_DESCRIPTION: This is a list of REs, which the block name must match, for statistics to be gathered. The following sample matches all blocks:
STATISTICS_BLOCKS_DESCRIPTION => [.*],
3-44
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
STATISTICS_OUTPUT_DIRECTORY: This is the name of the directory to which to write the block statistics files. This directory must exist and must have write privileges.
STATISTICS_OUTPUT_DIRECTORY => ../block_statistics
BLOCK_NAME: This is the name to use for each of the statistics blocks written to the LIF.
BLOCK_NAME => GATEWAY_BLOCK
OUTPUT_FILENAME_START: This is an optional text string to prepend to the block statistics filename.
OUTPUT_FILENAME_START => block_stats
A entire configuration can also be seen in the StatisticsConfig.pm file. A summary of the Block Statistics data is used in the notification event command sent to the notification engine, if Block Statistics gathering has been configured.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-45
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
60
Counter statistics are collected both from the raw input files and the LIF output. The counter statistics are output to a single LIF file for each Gateway run. The counter statistics are output, one counter per block, with the following information: The name of the counter The group (block name) to which it belongs Whether the counter was in the configured list for monitoring Whether the counter was present in the raw file Whether the counter was present in the LIF file
The counter statistics configuration entries are described in the next slides.
3-46
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
| Lotus software
STATISTICS_HEADER_MAPPINGS
61
STATISTICS_HEADER_ENTRIES: This is the list of header names and values to insert into the LIF header block.
STATISTICS_HEADER_ENTRIES => { GatewayID => Nortel, Network => GPRS CN, Version => 1.2, SenderType => HLR, SenderID => HLR-MSCW, },
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-47
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
STATISTICS_HEADER_MAPPINGS: This hash contains the names of the header counters to be mapped from header counters in the LIF block. Each entry is in the form <output_name> => <lif block name> where <output_name> is the name of the counter in the Counter Statistics file and <lif block name> is the name in the output LIF block.
STATISTICS_HEADER_MAPPINGS => { START_DATE => StartDate, START_TIME => StartTime, END_TIME => EndTime, SENDER_ID => MeasType, },
3-48
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
DISABLE_STATS OUTPUT_FILENAME_START
62
HEADER_INFO_FOR_STATS_FILE: This is an array containing the list of entries from STATISTICS_HEADER_ENTRIES and STATISTICS_HEADER_MAPPINGS that are to be used in creating the counter statistics LIF filename.
STATISTICS_OUTPUT_DIRECTORY: This is the path to the directory for the counter statistic files.
STATISTICS_OUTPUT_DIRECTORY => ../counter_stats
COUNTER_LIST: This is the list of counters that are specifically required to be checked for raw and LIF files. Any counter that is contained in this list will have IN_CONFIG set to TRUE in the output LIF. All other counters found that are not in this list will have IN_CONFIG set to FALSE.
COUNTER_LIST => [qw (VS.C7LKSYNU VS.LKFAIL VS.C7COV VS.C7ONSET1) ],
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-49
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
DISABLE_STATS: This is used to disable Counter_Statistics in the Gateway, if required. If set to TRUE, no counter statistics will be run. This may be done for performance reasons if there is a problem with backlogs or load.
DISABLE_STATS => 0,
3-50
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Student Exercise
IBM Software Group | Tivoli software
Student Exercise
Check the StatisticsConfig.pm snapshots in the next slide, and explain the expected Statistics behavior of your Framework at the blocks level. In your opinion, what can be done in this example to enhance the Statistics report capabilities of your Framework?
63
Student Exercise
64
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-51
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
For MPM
remsh servername -l metrica '/Interfaces/vendor_3.1/vstart/mpmalarm -u "Mon Sep 26 17:00:54 2005" -n "Alarm" -o "Gateway" -x "Ftp could not open connection to h1" -j "slade:Ericsson UTRAN Gateway" -s "major" -t "TmnUdp"' 2>/dev/null >/dev/null
65
Historically, it has not always been easy to monitor the activity of a Gateway, especially with multiple vendor Gateway installations. One of the reasons for this was that all activities that were reported to log and audit files, were not always accessible. The Gateway Framework has been updated to include configurable notification functionality that will send event and alarm notifications. These notifications should logically go to the application importing the output data from the Gateway. The current release of the Gateway has support for the following application notification Application Program Interfaces (APIs): Performance Alarm Server (PAS) for Network Performance Reporting (NPR): npralarm (supporting both event and alarm messages) Metrica Performance Manager (MPM) Notification Command Line Interface: mpmalarm (supporting both event and alarm messages). It can be also used as reference for PMW alarms. For Performance Manager for Wireless (PMW), this is reserved for future use.
The Notification Engine dispatches two types of activities from the Gateway Framework:
Alarm activities: Critical exit status and errors during processing of data Event activities: Statistical information with regards to blocks and files processed
3-52
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Alarm Notifications
Alarm notifications are sent for the following system errors: Engine module failure Disk space failures Failure to fork processes File transfer failures
Note that configuration failures will not generate alarm notification failures, because these are typically setup problems. See the some sample NPR and MPM alarm messages generated by notification functionality on the previous page; NotificationsConfig is reserved for future use in PMW.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-53
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Lotus software
-
remsh servername -l metrica '. /LOCAL/metrica/npr/setup.sh;npralarm target "TmnUdp" -objectClass "Gateway" -ObjectName "slade:Ericsson UTRAN Gateway" -name "Event" -text "File stats: Type statsfile, 10Aug2005 11:30, O/P file size total 30, Raw file size total 44, No files bad 0, No files OK 1" -severity "indeterminate"' 2>/dev/null >/dev/null
66
Event Notifications
Event notifications will be dispatched, if configured, for the following events: Engine file parsing completed Statistics of Blocks processed, if Block Statistics collection is configured Statistics of files processed, if File Statistics collection is configured. Note that the File Statistics summary rules (JOIN and ACCUMULATE) and functionality are not required because File Statistics event notifications are sent for each run of the Gateway. File statistical data entries used in notification event messages are as follows:
3-54
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
File type Output file size total Raw file size total No files bad 0 No files OK
A File Statistics notification event is dispatched for every File Statistics file type collected. Block Statistics message sends the total number of blocks and the total number of block counters. See the sample NPR event command messages generated by notification functionality on the previous page. Only one event and one alarm notification rule entry is allowed in the configuration. The type of rule hash entry is governed by NOTIFICATION_TYPE. This reflects the design where there is only one gathering source, each for all event and alarm notifications respectively. In the following slides you will see the standard notification entries to be included in the NotificationConfig.pm module.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-55
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
| Lotus software
67
RULE_TYPE: This is the rule type to dispatch messages. This must be the same as the name of the Perl module that implements the notification system. Note that this modularized approach also facilitates the inclusion of new Notification modules as required.
RULE_TYPE => NPR_Notification
RULE_DESC: This is a text string that will be logged to the audit file. This can be useful for tracing execution of the notification rule.
RULE_DESC => NPR Alarm notification,
NOTIFICATION_TYPE: This is the notification activity to dispatch. This is either Alarm or Event. Note that only one of each is allowed in the configuration.
NOTIFICATION_TYPE => Alarm
NOTIFICATION_TARGET: This is a text string of the target defined in the notification system.
NOTIFICATION_TARGET => TmnUdp
3-56
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
REMOTE_SHELL_COMMAND: This is the command line to run a remote shell on the remote server. If a remote command is not required, as is the case if the Gateway is sharing a host system with the service assurance application, then the REMOTE_SHELL_COMMAND value will reflect this.
REMOTE_SHELL_COMMAND => rsh server_name l metrica
SYSTEM_ENV_SETUP: These are environment setup scripts or commands for the notification system.
SYSTEM_ENV_SETUP => . /LOCAL/metrica/npr/setup.sh
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-57
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
RULE_TYPE => 'NPR_Notification', RULE_DESC => 'NPR Event Notification', NOTIFICATION_TYPE => 'Event', NOTIFICATION_COMMAND => 'npralarm', NOTIFICATION_TARGET => 'EventTmnUdp', REMOTE_SHELL_COMMAND => 'rsh servername -l metrica', SYSTEM_ENV_SETUP => '. /LOCAL/metrica/npr/setup.sh', },
68
A sample configuration of the Notification Engine for both Alarm and Event activities rules is shown in the slide.
3-58
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
MIN_FILES_PER_PROCESS
69
The Gateway can be configured for parallel processing. This allows a single Gateway to assign certain engine and Post-Parser rules across multiple processors. Configuring some specific Post-Parser rules (outlined on page 3-63) in the UserConfig.pm with these variables allow the Parallel Processing Before you consider configuring Engine or Post-Parser rules over multiple processors, the following points should be taken into account: Has the Gateway been analyzed to ensure it is configured for optimum performance? Ensure that the Gateway is as efficient as possible before considering splitting it over multiple processors. Are there multiple processors available to run the Gateway on? The rules should not be configured to run with more processes than there are processors on the network. Is there spare capacity that can be used? Ensure that the server capacity has not been exhausted already. Is there an advantage to be gained? The engine or Post-Parser rule should take at least 10 seconds to process its set of files before it is considered for multiple processor configuration. There is an overhead with spawning UNIX processes and subsequently managing them.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-59
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
There are a number of internal constraints on parallel processing: The PIF_Handler must be used so that PIF files are being read and written to disk, and can easily be shared between processes. The PIF_Cache in memory cannot be used. File_Statistics must not be configured.
Parallel processing is configured individually for each engine and Post-Parser rule. For example, if there were 100 files to be processed by a particular Post-Parser rule, these could be split over four processes, with each process handling 25 files. These processes run in parallel and are monitored by the Gateway. They all must be completed before the next rule is initiated.
3-60
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Example 2
NUMBER_OF_PROCESSES_DIST => { '00-02' => 1, '03-04' => 2, }
70
Parallel Processing is configured individually for each engine and Post-Parser rule. There are three configuration variables that control its use. NUMBER_OF_PROCESSES: This is the number of processes to split the processing of the rule over. Should not be more than the number of actual processors on the server.
NUMBER_OF_PROCESSES => 4
MIN_FILES_PER_PROCESS: This is the minimum number of files to assign per process. This value ensures that processes are only created, if there are sufficient files available. For example, for a rule processing very large files which can take 20 seconds to parse, it would be acceptable to assign one file per process. For other Post-Parser rules, each file can take such a short period that at least 20 files per process would need to be assigned.
MIN_FILES_PER_PROCESS => 20
NUMBER_OF_PROCESSES_DIST (Optional): This entry can be used to configure different process distributions over certain hourly periods. For example, it may be required that only two processes run at certain hours during backup periods to reduce the load. The following example specifies only one process from midnight to 2 a.m. and two processes from 3 a.m. to 4 a.m.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-61
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Some important points to be noticed are: All engine rules support parallel processing. Parallel processes cannot write to the same log or audit file; each parallel process will create its own log and audit file. The log and audit files will be the name of the base file suffixed by .<X> where <X> is the child number. The main Gateway process will continue to write to the main log and audit file.
For example, if the Gateway is using the files log and audit, and two processes are being created for parallel processing, then the following files will be created: log.0 log.1 audit.0 audit.1
3-62
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
MERGE_RECORDS
71
Certain rules cannot support parallel processing due to their functionality, requiring input of more than one file type. The Post-Parser rules that support parallel processing are shown in the slide and described in detail in the following unit.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-63
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Review Questions
IBM Software Group | Tivoli software
Review Questions
Describe in a paragraph how the Gateway Framework can be connected to a vendor Gateway.
72
3-64
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Student Exercise
IBM Software Group | Tivoli software
Student Exercise
Following the guidance of your instructor, you will be divided into groups and each group will be responsible for checking the following configuration files:
UserConfig.pm EngineConfig.pm TransferConfig.pm StatisticsConfig.pm
Write down any modifications that you believe will be significant to allow the Gateway operation. Discuss your opinions in class.
73
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
3-65
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Summary
IBM Software Group | Tivoli software
Summary
You should now be able to:
Perform basic Gateway Framework configuration. Start the Gateways using cron. Set up transfer configuration. Set up parser engine configuration. Set up Post-Parser configuration. Set up Statistics configuration. Configure Gateway notifications. Configure Gateways for parallel processing.
74
3-66
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
4-1
Introduction
This unit describes the standard Post-Parser rules supplied with the Gateway Framework. These rules are configured in UserConfig.pm. If they do not meet the requirements for the manipulation of output data, then they can be augmented by vendor specific Post-Parser rules. Each rule is accompanied by a sample of the input and output PIF, if applicable, to demonstrate its use. Where configuration is not straightforward or is in a complex structure, an example is included in the description.
Objectives
IBM Software Group | Tivoli software
Objectives
Upon completion of this unit, you will be able to:
Identify the standard Post-Parser rules. Explain the following Post-Parser rules:
ACCUMULATE AGGREGATE_LINE BATCHFILES CVAL_MANIP DATALINE_WHERE FILE_SPLIT INFOINSERT JOIN MERGE_RECORDS PERLIZE PIF_2_OUTPUT PIF_REMOVE SPLIT_RECORDS UNPEGGER
76
4-2
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
ACCUMULATE
IBM Software Group | Tivoli software
ACCUMULATE
(A)
##START|DATA_BLOCK CELL|COUNT1|COUNT2|COUNT3|L OAD|TRX DF0001|1|2|3|100|1 DF0001|1|2|3|10|2 DF0001|1|2|3|50|3 DF0001|1|2|3|90|4 DF0002|1|2|3|60|1 DF0002|1|2|3|0|2 DF0002|1|2|3|30|3 DF0002|1|2|3|20|4 ##END|DATA_BLOCK DF0002|0|4|8|60|4|12 ##END|ACCUM
(B)
77
Given the block of PIF data in (A), the output after processing by the ACCUMULATE rule is shown in (B). In this example the counters COUNT1, COUNT2, and COUNT3 are being accumulated using CELL value as the row key, with the maximum and minimum of the LOAD counter also being derived. Note in the output: The three counters COUNT1, COUNT2, and COUNT3 have been accumulated across the CELL. The minimum and maximum value have been placed in LOAD_MIN and LOAD_MAX respectively. The number of rows that matched the key to create the accumulated values is contained in ACCUM_NUM_IN_SUM.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-3
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
COUNTERS_TO_SORT_ON
78
This section describes the rule-specific configuration entries for the ACCUMULATE rule. It does not include the standard entries supported by every Post-Parser rule, already described in the last unit. The mandatory configuration entries are: OUTPUT_BLOCK_NAME: This is the block name to use in the output PIF or LIF. COUNTERS_TO_SORT_ON: This is the set of counters to use as the key for accumulation.
COUNTERS_TO_SORT_ON => [qw(CELL)],
4-4
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
79
NON_ADDITIVE_COUNTERS: Counters that are not to be accumulated, but pass through to the output as they are found. For example, date and time fields. REDUNDANT_DATA_COUNTERS: These are the data counters to remove from the output blocks. APPEND_STR: This is a string to append to the accumulated counters OLD_COUNTER_NAMES: This is a list of old counters to rename. NEW_COUNTER_NAMES must also be configured. NEW_COUNTER_NAMES: This is a list of the new counter names to replace the OLD_COUNTER_NAMES with. COUNTER_NULL_VALUE: This is a value to insert for the accumulated counters when there is an error calculation in their value. For example, when the values are non-numeric. MAXIMUM_COUNTERS: This is a list of counters for which to derive the maximum values.
MAXIMUM_COUNTERS => [qw(LOAD)],
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-5
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
MAXIMUM_APPEND_STR: This is the string to append to the MAXIMUM_COUNTERS to create the name of the new counter.
MAXIMUM_APPEND_STR => _MAX,
MINIMUM_COUNTERS: This is the list of counters from which the minimum values are derived.
MINIMUM_COUNTERS => [qw(LOAD)],
MINIMUM_APPEND_STR: This is the string to append to the MINIMUM_COUNTERs to create the name for the new counter containing the minimum value.
MINIMUM_APPEND_STR => _MIN,
4-6
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
See the sample configuration used in the sample application of the rule (see page 4-3).
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-7
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
AGGREGATE_LINE
IBM Software Group | Tivoli software
AGGREGATE_LINE
(A)
##START|DATA_BLOCK C_1_1|C_1_2|C_1_3|C_1_4|C_2 _1|C_1_5|C_2_2|C_2_3 20|20|10|30|20|30|10|20 10|0|40|20|0|0|0|10 20|20|30|30|10|0|0|10 10|30|40|0|20|0|10|20 30|0|10|0|20|30|10|20 ##END|DATA_BLOCK 110|20|20|10|30|20|30|10|20|50 70|10|0|40|20|0|0|0|10|10 100|20|20|30|30|10|0|0|10|20 80|10|30|40|0|20|0|10|20|50 70|30|0|10|0|20|30|10|20|50 ##END|DATA_BLOCK ##START|DATA_BLOCK C_1_Total|C_1_1|C_1_2|C_1_3|C_1_ 4|C_2_1|C_1_5|C_2_2|C_2_3|C_2_ Total
(B)
81
The AGGREGATE_LINE rule, similar to ACCUMULATE, accumulates a set of values. But rather than accumulate across a set of rows, the values are accumulated from a set of counters on a particular row. The set of counters for accumulation are matched through a series of regular expressions. All counters matching the regular expression are accumulated into a single value. To illustrate that, refer to the input PIF file in (A). After processing with AGGREGATE_LINE, the set of five counters C_1_1 to C_1_5, and the set of counters C_2_1 to C_2_3 are going to be totalled for each row of data into two new counters, C_1_Total and C_2_Total.
4-8
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
COUNTER_GROUPS
82
This section describes the rule-specific configuration entries for the AGGREGATE_LINE rule. It does not include the standard entries supported by every Post-Parser rule, already described in the last unit. The mandatory configuration entries for the AGGREGATE_LINE rule are: DEFAULT_NULL_VALUE: This is the value to insert for the new counter, if no counters in the data match it for accumulation. COUNTER_GROUPS: This is a hash containing a list of new counter names, each of which maps to an existing set of counters to be accumulated using a RE.
COUNTER_GROUPS => { C_1_Total => C_1_\d+, C_2_Total => C_2_\d+, },
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-9
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
| Lotus software
REDUNDANT_HEADER_COUNTERS
83
The optional configuration entries for the AGGREGATE_LINE rule are: COUNTER_NAME_TO_EXTRACT: As well as defining new counters to be created in COUNTER_GROUPS, this entry can be used to define a substring to extract from the counter name. This will be added to the front of the new COUNTER_GROUP counters created. OUTPUT_BLOCK_NAME: This is the name to use for the output block. REDUNDANT_DATA_COUNTERS: This is a list of data counters to remove from the output. REDUNDANT_HEADER_COUNTERS: This is a list of header counters to remove from the output.
4-10
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
84
See the sample configuration used in the sample application of the rule (see page 4-8).
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-11
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
BATCHFILES
IBM Software Group | Tivoli software
BATCHFILES
(A)
BSC-#-02Mar2004-#-1200-#-103-#-I.pif BSC-#-02Mar2004-#-1200-#-132-#-I.pif BSC-#-02Mar2004-#-1200-#-131-#-I.pif NSS-#-02Mar2004-#-1200-#-172-#-I.pif NSS-#-02Mar2004-#-1200-#-131-#-I.pif NSS-#-02Mar2004-#-1200-#-173-#-I.pif
BATCHED-#-BSC-#-02Mar2004-#-1200-#-1-#-BF.lif BATCHED-#-NSS-#-02Mar2004-#-1200-#-1-#-BF.lif
85
The BATCHFILES rule groups, or joins, a number of files together to create one large output file, usually a LIF. The batching of smaller PIF objects is usually performed to reduce the number of files being presented to the PM system for loading. See the slide for reference. Given the list of PIF files in (A), you want to create the large LIF files in the output (B), where the BSC and NSS files will be batched together to create a single BSC and NSS file for loading. The files are prefixed with the BATCHED string, followed by the matched pattern from the INPUT_FILE_DESCRIPTION, in this case the type and date time, for example, BSC-#02Mar2004-#-1200.
4-12
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
This section describes the rule-specific configuration entries for the BATCHFILES rule. It does not include the standard entries supported by every Post-Parser rule, already described in the last unit. The optional configuration entries for the BATCHFILES rule are: OUTPUT_BLOCK_NAME: This is the block name to use in the output data. REDUNDANT_HEADER_COUNTERS: This is a list of header counters to remove in the output data. REDUNDANT_DATA_COUNTERS: This is a list of data counters to remove in the output data. OUTPUT_FILENAME_START: This is a string to append to the filename HOURS_TO_WAIT_FOR_PARTNER_FILES: This is the number of hours to wait after the last file, before the files present are processed without their partner files. If set to -1, it will not wait for partner files. QUOTE_INPUT_PIFS: If set to TRUE, the names of the PIFs used to create the output performance LIF will be included as a comment at the top of the file.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-13
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
87
See the sample configuration used in the sample application of the rule (see page 4-12).
4-14
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
CVAL_MANIP
IBM Software Group | Tivoli software
CVAL_MANIP
(A)
##START|BLOCK_DATA
C_3|C_5|C_1|C_2 IBM Software Group | Lotus software 3|5|"1 10"|2 3|5|"2 14"|2 3|5|6|2 3|5|2|2 3|5|8|2 ##END|BLOCK_DATA
##END|BLOCK_DATA
88
The CVAL_MANIP Post-Parser rule provides the functionality to manipulate counter values. Examples of its use are: Wrapping a counter value in quotes Removing a substring from a counter value Replacing a counter value Rearranging a counter value
See the slide, with input PIF file in (A), and the results of a CVAL_MANIP usage in (B). In this slide, the counter value manipulation will be performed on counter C_1. The following operations will be performed on this counter: Any values prefixed by KEY- will have that portion of the value removed. Any values with one or more spaces will be quoted and the value replaced with one space.
Notice also that C_4, considered a redundant counter in this example, that is, no incrementing occurs, has been fully erased from output (B).
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-15
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
MATCH
IBM PATTERN
89
This section describes the rule-specific configuration entries for the CVAL_MANIP rule. It does not include the standard entries supported by every Post-Parser rule, already described in the last unit. The mandatory configuration entries for the CVAL_MANIP rule are: CNAME_MANIP: This is the name of the counter to be manipulated.
CNAME_MANIP => C_1,
MATCH: This is a list of regular expressions (in order of preference) that the counter values are expected to match in order for the values to be manipulated. Each regular expression should contain at least one pattern to be extracted if the value is to be manipulated.
MATCH => [^(KEY)\-(\d+),^(\d+)\s+(\d+)],
PATTERN: This is a list of patterns that define how the counter value will be manipulated. There is one-to-one correspondence between the position of the regular expression in list MATCH and the patterns in PATTERN. Any occurrence of $1, $2, ... in PATTERN will be substituted by the values of the tagged expressions matched by the corresponding regular expression in MATCH. Hence, the list MATCH and PATTERN must have the same number of elements.
PATTERN => [$2, $1 $2]
4-16
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
OUTPUT_DIR
| Lotus software
90
The optional configuration entries for the CVAL_MANIP rule are: OUTPUT_BLOCK_NAME: This is the block name to use in the output data. OUTPUT_DIR: This is the name of an alternate directory to which the loader files can be written for this rule. If not configured, then LIFs will be written to the configured parser output directory. REDUNDANT_DATA_COUNTERS: This is a list of counters to remove from the output data. NON_MATCH_RECORD: This value determines whether or not to output a PIF record whose counter value fails to match any of the patterns. If set to TRUE, it will discard the row.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-17
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
CVAL_MANIP_Example Configuration
IBM Software Group | Tivoli software
91
See the sample configuration used in the sample application of the rule (see page 4-15).
4-18
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
DATALINE_WHERE
IBM Software Group | Tivoli software
DATALINE_WHERE
(A)
##START|DATA_BLOCK
92
The DATALINE_WHERE rule is used to either remove, or to keep PIF data rows based on the counter values. In the slide, where (A) is the PIF input to the DATALINE_WHERE rule, output (B) is obtained filtering the rows which match the condition any row with KEY value 1-5 AND with a C_1 value not equal to 1 or 3 will be kept. Notice that only one data row matches both criteria.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-19
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
93
This section describes the rule-specific configuration entries for the DATALINE_WHERE rule. It does not include the standard entries supported by every Post-Parser rule, already described in the last unit. The mandatory entry for the DATALINE_WHERE rule is: COUNTER_NAMES: This is the entry is mandatory and is configured as an array of anonymous hashes. Each hash entry contains: COUNTER_NAME: This is the name of the counter to be checked. KEEP_WHERE (optional): This is a list of regular expressions. The counter value must match ONE of these REs for the data line to be kept. REMOVE_WHERE (optional): This is a list of regular expressions. If the value counter matches ONE of these REs, then it will be removed. Either KEEP_WHERE or REMOVE_WHERE must be configured. If both are configured, then any data rows which pass the KEEP_WHERE list will be checked against the REMOVE_WHERE list.
COUNTER_NAMES => [ { COUNTER_NAME => KEY, KEEP_WHERE => [^[1-5]], REMOVE_WHERE => [],
4-20
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-21
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
NEW_COUNTER_VALUE FILENAME_ADDITION
94
The optional configuration entries for the DATALINE_WHERE rule are: OUTPUT_BLOCK_NAME: This is the block name to use in the output files. REDUNDANT_DATA_COUNTERS: This is a list of counters to remove from the output data. ADD_NEW_COUNTER: This is the name of a new counter to be added to the output data. NEW_COUNTER_VALUE: This is the value of the new counter. FILENAME_ADDITION: This is a string inserted into the output filename before the -#-DW sequence.
4-22
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
95
See the sample configuration used in the sample application of the rule (see page 4-19).
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-23
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
FILE_SPLIT
IBM Software Group | Tivoli software
FILE_SPLIT
(A)
##START|FS
(B)
##START|BLOCK C_3|C_4|C_5|TIME|KEY|C_1|C _2 2|6|9|00:00|KEY_0|0|1 7|5|2|01:00|KEY_1|1|5 1|4|5|02:00|KEY_2|0|4 6|7|9|00:00|KEY_0|2|5 4|4|6|01:00|KEY_1|5|1 2|8|3|02:00|KEY_2|3|6 3|6|3|00:00|KEY_0|0|6 4|0|7|01:00|KEY_1|4|3 ##END|BLOCK
(D)
96
The FILE_SPLIT rule parses PIF files into smaller files based a split key made up of one or more counter values. Examine the slide, with (A) being the Input PIF file. The FILE_SPLIT rule will split (A) into the smaller files (B), (C) and (D). In this situation, this PIF data is split on the TIME and KEY counters. Each data row with a different file split key from these counter values will be output to a separate PIF. The split key will also be used in the PIF filename to ensure the split filenames are unique. In this case there are three different keys producing three PIFs: (B) File fs_in-#-00:00-#-0-#-FS.pif (C) File fs_in-#-01:00-#-1-#-FS.pif (D) File fs_in-#-02:00-#-2-#-FS.pif
4-24
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
SPLIT_COUNTERS_ORDER
97
This section describes the rule-specific configuration entries for the FILE_SPLIT rule. It does not include the standard entries supported by every Post-Parser rule, already described in the last unit. The mandatory entries for the FILE_SPLIT rule are: COUNTERS_USED_TO_SPLIT_FILE: This is an expression containing the information on how to split the file. The key is the counter names to split on and the value is a regular expression describing the portion of the counter value to extract as a key.
COUNTERS_USED_TO_SPLIT_FILE => { TIME => '(.*)', KEY => '^KEY_(\d+)', },
SPLIT_COUNTERS_ORDER: This is the order of the counter keys used to split the file in the new filename.
SPLIT_COUNTERS_ORDER => [ qw(TIME KEY) ],
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-25
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
REDUNDANT_HEADER_COUNTERS
| Lotus software
98
The optional configuration entries for the FILE_SPLIT rule are: OUTPUT_BLOCK_NAME: The block name to use in the output files. REDUNDANT_HEADER_COUNTERS: These are omitted from the output data. REDUNDANT_DATA_COUNTERS: These are omitted from the output data.
4-26
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
See the sample configuration used in the sample application of the rule (see page 4-24).
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-27
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
INFOINSERT
IBM Software Group | Tivoli software
INFOINSERT
##START|BLOCK CELL|INFO_KEY|TRX 10-0|1|10-10-49 10-0|2|10-10-50 10-2|3|10-10-60 10-3|4|10-10-69 10-3|5|10-10-79 10-3|6|10-10-50 10-0|7|10-10-22 10-2|8|10-10-80 10-3|9|10-10-77 10-7|10|10-10-27
(C)
(A)
##END|BLOCK
(B)
##END|BLOCK
100
The INFOINSERT rule inserts counter data from a secondary information file into a primary data file. It is typically used to insert hierarchy data from a configuration file into a main file, based on a counter key. These counter keys may be made up of one or more counters, with both header and data counters configurable for both the primary and secondary files. The INFOINSERT rule requires at least two input files for processing. The file from where data will be extracted (A, in the example) and the file where data will be inserted (B). In this example the key matched between the files is the OBJ_ID from the primary file and INFO_KEY from the secondary file. The counters CELL and TRX are being inserted from the secondary file. This creates the output (C).
4-28
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
| Lotus software
HEADER_NAMES-USED_TO_ID_INFORMATION
NAMES_USED_TO_ID_INFORMATION
101
This section describes the rule-specific configuration entries for the INFOINSERT rule. It does not include the standard entries supported by every Post-Parser rule, already described in the last unit. The mandatory entries for the INFOINSERT rule are: HEADER_NAMES_USED_TO_ID_DATA_RECORD: This is a list of header names that form the first part of a key that is used to identify which record of secondary information should be insert in this data record.
HEADER_NAMES_USED_TO_ID_DATA_RECORD => [ ],
NAMES_USED_TO_ID_DATA_RECORD: This is a list of counter names that are used to construct the second part of an identifier that is used to choose which record of secondary information should be included with each data record.
NAMES_USED_TO_ID_DATA_RECORD => [OBJ_ID],
INFO_FILE_DESCRIPTION: This is a regular expression, or a list of regular expressions describing the names of the secondary files that contain the information that is to be substituted into the data lines of the files that are described in the option INPUT_FILE_DESCRIPTION.
INFO_FILE_DESCRIPTION => [info.pif],
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-29
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
HEADER_NAMES_USED_TO_ID_INFORMATION: This is a list of header counter names that form the first part of the key that is used to create the unique key to identify the secondary data for insertion.
HEADER_NAMES_USED_TO_ID_INFORMATION => [ ],
NAMES_USED_TO_ID_INFORMATION: This is a list of counter names used to construct a unique identifier for the records of data in the secondary information file.
NAMES_USED_TO_ID_INFORMATION => [INFO_KEY],
4-30
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
| Lotus software
102
The optional entries for the INFOINSERT rule are: ONLY_INSERT: This list can be used to configure the list of counter names that are required for insertion, if the full set of data from the information file is not required.
ONLY_INSERT => [CELL, TRX],
WRITE_DATA_LINE: This option controls the action of this rule when there is no information to substitute for a line of data, that is, the key from the main file is not found in the secondary file. If set to TRUE, and there is no data to substitute, the data row will be output anyway, with NULL values inserted for the secondary keys. If set to FALSE the data row will not be output. OUTPUT_BLOCK_NAME: This is the name that should be used for the section name in the loader file. OUTPUT_FILENAME_START: This is a prefix that the output file name will start with. REDUNDANT_DATA_COUNTERS: This is a list of counters that should be deleted from the resultant output.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-31
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
INFO_FILES_STORAGE_DIR: This is an optional scalar entry containing a directory name where information files can be stored. Information files not stored are deleted, as will be the case if INFO_FILES_STORAGE_DIR is not set in the INFOINSERT configuration, or is set to zero (in non-debug mode). REMOVE_INFO_FILES: By default the information files are kept for the run. In certain situations, where the information files are being created for each Gateway iteration, this is not necessary. If this option is set to TRUE, the information files will be deleted.
4-32
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
HEADER_NAMES_USED_TO_ID_DATA_RECORD => [ ], => ['OBJ_ID'], => ['info.pif'], => ['INFO_KEY'], => 'INS_BLOCK', => 'II',
103
See the sample configuration used in the sample application of the rule (see page 4-28).
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-33
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
JOIN
IBM Software Group | Tivoli software
JOIN
##START|CELL_HO CHO_3|CELL_ID|BTS_ID|CHO_1|CHO_2 3|10-10-1|10-1|1|2
(C)
##START|CELLDATA CHO_3|CELL_ID|CTRF_1|BTS_ID|CTRF_2|CHO_1 |CTRF_3|CHO_2 3|10-10-6|25|10-0|78|1|103|2
(A)
3|10-10-3|35|10-0|193|1|7|2 IBM Software Group | Lotus software 3|10-10-4|90|10-1|158|1|82|2 3|10-10-1|32|10-1|91|1|118|2 3|10-10-5|38|10-2|47|1|50|2 3|10-10-2|5|10-2|123|1|290|2 ##END|CELLDATA
##START|CELL_TRAFFIC
(B)
104
The JOIN rule is used to connect rows from multiple files into a single file based on the following: Pattern matching in the file name Counter key matching within each file
The JOIN rule produces one larger file, with a single data row containing the data rows from the individual files in the output. The example requires a number of input files to show the full usage. Two input files are configured:
File CELL-HO-#-20Mar2004-#-01:00-#-I.pif (A) File CELL-TRAFFIC-#-20Mar2004-#-01:00-#-I.pif (B).
These two files are joined based on the date time pattern in the filename, and on the counter keys BTS_ID and CELL_ID. This creates one output file CELLDATA-#-1-#-J.pif with the joined contents (C). The joined file contains the data rows from each PIF joined to create a single row containing both the handover and traffic counters for each CELL.
4-34
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
OUTPUT_BLOCK_NAME
105
This section describes the rule-specific configuration entries for the JOIN rule. It does not include the standard entries supported by every Post-Parser rule, already described in the last unit. The mandatory entries for the JOIN rule are: COUNTERS_TO_JOIN_ON: This is the list of counters to join the input files on.
COUNTERS_TO_JOIN_ON => [qw(CELL_ID BTS_ID) ],
OUTPUT_BLOCK_NAME: This is the name of the new data block in the output file.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-35
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
The optional configuration entries for the JOIN rule are: REDUNDANT_HEADER_COUNTERS: This is a list of counters to remove from the header data block. REDUNDANT_DATA_COUNTERS: This is a list of counters to remove from the data block. OUTPUT_FILENAME_START: This is a prefix to use on the output file name. HEADER_COUNTERS_TO_USE_IN_OUTPUT_FILENAME: This is a list of counter names to use to construct the output file name. HOURS_TO_WAIT_FOR_PARTNER_FILES: This is the amount of time for a PIF file to wait for its partner file before it can be joined. This entry should be removed from the configuration or set to -1 if there is no partner file to wait for. TIME_JOINFILE_PRODUCED: If set to TRUE, the rule will add time and date to the output joined file. The time and date is in the following format: <DAY><DATE><MON><YEAR>_HH:MM:SS QUOTE_INPUT_PIFS: If set, the rule will add the names of the input PIFs to the output LIF. This can be useful when trying to debug the joining of a large number of files or a complex rule.
Copyright IBM Corp. 2007
4-36
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
107
See the sample configuration used in the sample application of the rule (see page 4-34).
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-37
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
MERGE_RECORDS
IBM Software Group | Tivoli software
MERGE_RECORDS
(A)
##START|DATA_BLOCK OBJ_KEY|C_3|PEAK_TYPE|PEAK|C_1|C1 0_Total|C_2 10-10-1|3|TRAF|t200|1|1100|2 10-10-2|3|TRAF|t300|1|1200|2 10-10-3|3|TRAF|t400|1|1300|2 ##START|NEW_BLOCK OBJ_KEY|C_3|PEAK_TYPE|NUM_MERGED|TCPU_PE 10-10-0|3|TRAF|2|p1100|t600|1|1000|2 10-10-1|3|TRAF|2|p700|t200|1|1100|2 10-10-2|3|TRAF|2|p800|t300|1|1200|2 10-10-3|3|TRAF|2|p900|t400|1|1300|2 10-10-4|3|TRAF|2|p1000|t500|1|1400|2 ##END|NEW_BLOCK
(B)
AK|TTRAF_PEAK|C_1|C10_Total|C_2 IBM Software Group | Lotus software 10-10-4|3|TRAF|t500|1|1400|2 10-10-0|3|TRAF|t600|1|1000|2 10-10-1|3|CPU|p700|1|1100|2 10-10-2|3|CPU|p800|1|1200|2 10-10-3|3|CPU|p900|1|1300|2 10-10-4|3|CPU|p1000|1|1400|2 10-10-0|3|CPU|p1100|1|1000|2 ##END|DATA_BLOCK
108
The MERGE_RECORDS rule merges PIF data records within one PIF file based on a counter key. As each row contains the same counter names, the rule can be configured either to insert all values of a counter in the output row, or just a single value. In the example, given the input PIF data (A) you want to have the PIF data rows to be joined on the counter OBJ_KEY. All counters being merged onto the same row have the same value except for PEAK, which needs to be preserved from each row. This produces the output (B) , where: The counters C_1, C_2, C_3, and C10_Total are merged into a single value. A new counter NUM_MERGED is added which counts the number of rows merged to create the new line. The PEAK value from each row is preserved. The rule creates two new counters, TCPU_PEAK and TTRAF_PEAK, which have the counter value from each of the merged rows. The counter PEAK_TYPE has been as a grouping key to identify the different counter values from each merged row.
4-38
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
109
This section describes the rule-specific configuration entries for the MERGE_RECORDS rule. It does not include the standard entries supported by every Post-Parser rule, already described in the last unit. The mandatory entries for the MERGE_RECORDS rule is: COUNTERS_TO_SORT_ON: This is a list of counters for the merge key. All rows with the same key will be merged into one output row.
COUNTERS_TO_SORT_ON => [OBJ_KEY],
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-39
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
REDUNDANT_DATA_COUNTERS
GROUP_KEY
MANIP_ONLY
110
The optional configuration entries for the MERGE_RECORDS rule are: OUTPUT_BLOCK_NAME: This is the block name used in the output files. REDUNDANT_DATA_COUNTERS: This is a list of counters to remove from the output data. GROUP_KEY: This is a counter name whose values within a collective set of data to be merged can be regarded as a key. The values will be used to prefix the counter names of records to be merged. If this option is not configured, then a dummy prefix is used (T0, T1 ..).
GROUP_KEY => PEAK_TYPE,
MANIP_ONLY: This is a list of counter names whose values from all rows are to be preserved in the output data. If option is not configured, then all counter names other than those listed in option COUNTERS_TO_SORT_ON and REDUNDANT_DATA_COUNTERS (if configured) will be manipulated.
MANIP_ONLY => [PEAK]
4-40
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
111
See the sample configuration used in the sample application of the rule (see page 4-38).
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-41
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
PERLIZE
IBM Software Group | Tivoli software
PERLIZE
(A) ##START|HEADER HC_1|HC_2|HC_3|DATETIME 0|1|15|20Mar2004_12:00 ##END|HEADER ##START|DATA_BLOCK C_3|C_1|C_2 19|4|6 14|3|19 22|0|3 25|5|16 0|3|10 ##END|DATA_BLOCK ##START|HEADER DATE|HC_1|HC_2|HC_3|TIME 20Mar2004|0|1|15|12:00 ##END|HEADER (B)
##START|DATA_BLOCK IBM Software Group | Lotus software C_ACCUM|C_3|C_1|C_2 29|19|4|6 36|14|3|19 25|22|0|3 46|25|5|16 13|0|3|10 ##END|DATA_BLOCK
112
The PERLIZE rule, unlike other rules, has complete flexibility as to how it is used. The functions for manipulating the header or counter data are configured in the Post-Parser configuration file. The rule passes out each row of header and counter data to separate functions in the Post-Parser configuration. In other words, PERLIZE allows you to input direct Perl programming lines into the Gateway Framework. Here the data, both header and counter rows, can be manipulated in many ways including: Deriving new counters from existing values, including using mathematical operations Adding new counters Renaming counters Removing counters Filtering files through the header fields, and counter rows individually
In the slide, given the input data (A) you will have the PERLIZE rule configured to call out to the following two subroutines:
4-42
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
One subroutine manipulates the header data, and creates two new counters, DATE and TIME, from the current DATETIME counter. One subroutine manipulates the counter data, and creates a new counter, C_ACCUM, which is the sum of all counter values on each row.
This produces the following output (B) as shown in the slide on the previous page.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-43
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
113
This section describes the rule-specific configuration entries for the PERLIZE rule. It does not include the standard entries supported by every Post-Parser rule, already described in the last unit. The mandatory configuration entries for the PERLIZE rule is: FILENAME_SUFFIX: This is a suffix string to append to the output file. This should reflect the operation performed by PERLIZE.
4-44
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Lotus software
114
The optional configuration entries for the PERLIZE rule are: OUTPUT_BLOCK_NAME: This is the block name to use in the output data. OUTPUT_DIR: This is an alternative directory to write the performance LIF data to. REDUNDANT_HEADER_COUNTERS: This is a list of header counters omitted from the output data. REDUNDANT_DATA_COUNTERS: This is a list of data counters to omitted from the output data. HEADER_COUNTERS_OP: This is a subroutine to process the header block. This subroutine is passed a reference to a hash containing the header names and values. If the subroutine returns non 0 then the file is discarded. The following example is creating separate date and time counters from one combined header counter, DATETIME:
HEADER_COUNTERS_OP => sub { my $h_ref = shift; # split date/time into 2 separate counters my @data = split("_",$h_ref->{DATETIME}); $h_ref->{"DATE"} = $data[0];
Copyright IBM Corp. 2007 IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-45
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
4-46
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
115
DATA_COUNTERS_OP: This is a subroutine to process each PIF data row which is passed out as a hash reference. If the subroutine returns a non-zero value, then the row will be discarded. In the following example the total of all current counters is being output in a new counter C_ACCUM.
DATA_COUNTERS_OP => sub { my $c_ref = shift; # create a new counter my $accum = 0; foreach (values %$c_ref){ $accum += $_; } $c_ref->{C_ACCUM} = $accum; # keep row return 0; },
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-47
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
116
See the sample configuration used in the sample application of the rule (see page 4-42).
4-48
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
PIF_2_OUTPUT
IBM Software Group | Tivoli software
PIF_2_OUTPUT
(A)
#%npr # { { BSC 10-20-30 DATE 20Mar2004 STARTTIME 12:00 ENDTIME 12:15 H_1 2 H_2 4 OUT_BLOCK {
(B)
##START|BLOCK
117
The PIF_2_OUTPUT rule converts PIF based data to the final output format. It has two main functions: Configuring any output format that supports the LIF_Writer interface. The Gateway Framework contains standard CSV_Writer and XML_Writer modules, as output formats. Sorting header and data counters for data output. This is useful during development and installation, to track the values of specific counter names in the output data.
In the example, given the input PIF (A) you will have the PIF_2_OUTPUT rule run on this data to: Output the data using the LIF_Writer, therefore, the output will be LIF format. Output the header counters in BSC, DATE, STARTTIME order. Other counters not in this list will be sorted lexically. Delete the header counter H_3 Delete the data counter C_1
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-49
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
| Lotus software
This section describes the rule specific configuration entries for the PIF_2_OUTPUT rule. It does not include the standard entries supported by every Post-Parser rule, already described in the last unit. There are no mandatory entries other than those that apply to all Post-Parser rules. The optional entries for the PIF_2_OUTPUT rule are: OUTPUT_FORMAT: This is the module used to write out the performance data. This can be set to LIF_Writer, XML_Writer, CSV_Writer, or Prospect_Writer.
OUTPUT_FORMAT => LIF_Writer
REDUNDANT_HEADER_COUNTERS: This is a list of header counters to remove from the output data. REDUNDANT_DATA_COUNTERS: This is a list of data counters to remove from the output data. OUTPUT_BLOCK_NAME: This is the block name used in the output data. OUTPUT_FILENAME_START: This is a string to prepend to the output filename. OUTPUT_DIR: This is the directory to output the performance data to, if it is not to be written to the output directory configured for the Gateway.
Copyright IBM Corp. 2007
4-50
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
HEADER_COUNTERS_ORDER
| Lotus software
OUTPUT_RECORD_KEY_DELIMITER
119
NEW_HEADER_COUNTERS: This is a hash, containing a set of names and values of new counters to be output in the header. HEADER_COUNTERS_ORDER: This is the order in which to output the header counters. If not set no sorting is performed. Only list the counters that are required to be sorted and appear at the start of the output block. All other counters in the output data will be sorted lexically.
HEADER_COUNTERS_ORDER => [ qw(BSC DATE STARTTIME) ],
OUTPUT_RECORD_KEY: This is the list of data counters to form the record key as the first column of the output block records.
OUTPUT_RECORD_KEY => [ qw (ID_1 ID_2) ],
OUTPUT_RECORD_KEY_DELIMITER: This is a string to separate the counter names for the output record key. If not set, the default -#- is used.
OUTPUT_RECORD_KEY_DELIMITER => -,
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-51
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
120
See the sample configuration used in the sample application of the rule (see page 4-49).
4-52
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
PIF_REMOVE
IBM Software Group | Tivoli software
PIF_REMOVE
Removes PIF objects during Post-Parser stage.
Useful to reduce the complexity of configuring subsequent rules. Also reduces the memory and disk space requirements by freeing IBM Software Group | Lotus software up space before further Post-Parser rules run.
121
No examples are applicable for this rule, since all files that match the INPUT_FILE_DESCRIPTION of the rule will be deleted. In addition, there are no non-standard entries for this Post-Parser rule. In the example configuration on the next slide all BSC files with the suffix A are being deleted.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-53
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
{ RULE_TYPE => 'PIF_REMOVE', RULE_DESC => 'Remove all accumulated BSC files', IBM Software Group | Lotus
software
122
4-54
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
SPLIT_RECORDS
IBM Software Group | Tivoli software
SPLIT_RECORDS
(A)
14|10|SV_18|CV_11|12
C_3|COUNTER_ID|KEY|C_2
123
The SPLIT_RECORDS rule parses a single PIF data row into multiple rows, based on different counter names. Counters that are being used to split the data will have a new value in each new data rows. In the example, given the input PIF data (A), the rule will perform the following: Split each row into two rows, using the keys SV_KEY and CV_KEY. The type used to split the row will be placed in a new counter COUNTER_ID. The value that was contained in SV_KEY or CV_KEY will be output in a new counter KEY. The counters C_2 and C_3 will be written out in each data row.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-55
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
NEW_CNAMES
124
This section describes the rule-specific configuration entries for the SPLIT_RECORDS rule. It does not include the standard entries supported by every Post-Parser rule, already described in the last unit. The mandatory configuration entries for the SPLIT_RECORDS rule are: SPLIT_CNAMES: The key of the hash will be a unique string that will identify the new split record. The new split record will report values for the counter names list in its array. The size of the array can vary, but should not exceed the number of elements configured for the array NEW_CNAMES. The counter names listed in different arrays do not need to be mutually exclusive. Unless the options ONLY_INSERT or REDUNDANT_DATA_COUNTERS are configured, any counter names that are not listed in the arrays are automatically reported in each of the split records.
SPLIT_CNAMES => { T_CV => [CV_KEY], T_SV => [SV_KEY], },
NEW_CNAMES: This is an array of new counter names that will apply to the counters that have been split in the record.
NEW_CNAMES => [KEY],
4-56
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
REDUNDANT_DATA_COUNTERS
NEW_REC_ID_NAME
125
The optional configuration entries for the SPLIT_RECORDS rule are: OUTPUT_BLOCK_NAME: This is a replacement data block name used in the output files. OUTPUT_DIR: This is a full path to an alternative directory to which the output files will be written. ONLY_INSERT: This is a list of other counter names to be reported in every record. If not set, nor REDUNDANT_DATA_COUNTERS is set, no such counter names values will be inserted.
ONLY_INSERT => [C_3, C_2],
REDUNDANT_DATA_COUNTERS: This is a list of counters that should be deleted from the resulting output. If set, then all other counter names reported and not specified in SPLIT_CNAMES will be reported. If both REDUNDANT_DATA_COUNTERS and ONLY_INSERT, then ONLY_INSERT takes precedence. If neither is set, then no counter names will be inserted. NEW_REC_ID_NAME: This defines the new counter name in the block. It reports as its values the keys of the hash SPLIT_CNAMES.
NEW_REC_ID_NAME => COUNTER_ID,
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-57
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
See the sample configuration used in the sample application of the rule (see page 4-55).
4-58
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
UNPEGGER
IBM Software Group | Tivoli software
UNPEGGER
## IBM Parser Intermediate File ##START|HEADER
(A)
## IBM Parser Intermediate File ##START|HEADER
(B)
H_4|STARTTIME|DATE|ManagedElement|H_ 1|H_2|DURATION|H_3 4|14:15|02Mar2004|GGSN-R1D9|1|2|15|3 ##END|HEADER ##START|GGSN_BLOCK GgsnFunction|VS.IPSec.IncDataOct|VS. IPSec.DiscDataPkt|GiIsp|VS.IPSec. IncDataPkt|VS.IPSec.OutDataOct 0|961|2034|4|9995|5305 0|14312|1621|7|14441|2846 0|1455|7127|12|6533|4730 0|12076|4191|19|5113|14295 0|11901|9860|28|1607|3949 ##END|GGSN_BLOCK
H_4|STARTTIME|DATE|ManagedElement|H_1|DURAT ION|H_2|H_3 4|14:30|02Mar2004|GGSN-R1D9|1|15|2|3 ##END|HEADER ##START|GGSN_BLOCK GgsnFunction|VS.IPSec.IncDataOct|VS.IPSec.D IPSec.OutDataOct 0|18284|29817|4|29650|25838 0|15924|24286|7|28204|18781 0|24969|27115|12|28747|28223 0|17761|23425|19|28908|15957 0|21861|19129|28|17700|17415 ##END|GGSN_BLOCK
127
The UNPEGGER rule is used to derive unpegged values for pegged counter types. Pegged counters are defined as those whose value constantly increases up to a set value, until it rolls back to 0. A single value of a pegged counter contains no information on system performance. At least two values are required to derive the change in value over a period. For example: At 15:00 counter pdpContexts had a pegged value of 13456. At 15:15 counter pdpContexts had a pegged value of 14456. At 15:30 counter pdpContexts had a pegged value of 20000.
Therefore the unpegged value for this period at 15:15 is 1000, and at 15:30 the unpegged value is 5544. These types of counters are usually found in IP-based networks, and are either 32 or 64 bit in size. The application of this rule is quite complicated. The example described here, only gives the simplest application for the usage of the rule.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-59
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
the following processing is performed by the rule as described in the next slide:
UNPEGGER
(C)
##START|GGSN_BLOCK
128
Each GGSN pattern in the file name, GGSN-#-GGSN-R1D9 is processed as a group. Files with the GGSN pattern are then processed within the date time key, in this case 02Mar2004-#-14:15 followed by 02Mar2004-#-14:30. Since the file 14:15 has no previous file, no unpegged output is produced, the file is saved for the next iteration. The counter values in the 14:30 file are unpegged against the previous file 14:15. The DATE, STARTTIME, and DURATION counters in the header are used to derive the time difference between the files. The counters GgsnFunction and GiIsp are used as data row keys to match the related data rows between each PIF file. The VS.IPSec* counters are being unpegged. They are all 32-bit counters. The output PIF file will be prefixed with UNPEG.
4-60
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
This produces the output file (C) , UNPEG-#-GGSN-#-GGSN-R1D9-#-02Mar2004-#14:30-#-I.pif: This is the unpegged data. For example in the first row (GgsnFunction=0, GiIsp=4), the difference in value for counter VS.IPSec.IncDataOct from 14:15 to 14:30 is 18284 (from 14:30 file) less 961 (from 14:15 file) = 17323 The rule can equally be configured to unpeg data for multiple date or time periods in a single PIF file. The rule can also handle a backlog of files, such as multiple files being sorted in date and time order, and then processed in this order.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-61
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
129
Unpegging of valid data must be based on valid date, time, and duration calculations. The UNPEGGER rule supports a number of formats for the date and time in the pegged filename, as extracted using INPUT_FILE_DATETIME_KEY. These formats are: <day><month><year>[DATE SEP]<hour>[HOUR SEP]<minute> Where: Day: Is a 2-digit value Month: Is a 2-digit value (01, 02, and so forth), or a 3-letter prefix, (Jan, Feb, and so forth Year: Is a 4-digit value
[DATE SEP]: Is an optional date separator. Acceptable values for the separator are -, _, ., and -#-. Hour: This is in a 24-hour format Minute: Is a 2-digit value
4-62
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Examples of valid date and time values are: 02Mar2004-#-11:00 02032004_11.00 02Mar20041100
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-63
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
UNPEG_FILE_TYPE
KEEP_RAW_GRANULARITY
130
This section describes the rule-specific configuration entries for the UNPEGGER rule. It does not include the standard entries supported by every Post-Parser rule, already described in the last unit. The mandatory configuration entries for the UNPEGGER rule are: INPUT_FILE_DESCRIPTION: This is a list of pegged files to be processed by this rule. The element ID and key must be extracted in the configuration. This ensures that the correct element is unpegged against its previous file. In the following example the element type, GGSN, along with the element ID is being used as the pattern to match.
INPUT_FILE_DESCRIPTION => ^(GGSN-#-\w+\-\w+).*\.pif,
INPUT_FILE_DATETIME_KEY: This is a regular expression matching the input file name and extracting the date and time key from the filename. This is required to sort the input files in the correct order, so that the oldest input file will be processed first. INPUT_FILE_DATETIME_KEY =>
^.*-#-(\d{2}\w{3}\d{4}-#-\d{2}:\d{2})-#-I\.pif,
4-64
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
PEG_FILENAME_PREFIX: This is the prefix used to generate the previous pegged PIF file. After the current file has been unpegged, it must be saved as the previous file. The values in this file will be used as the previous values for the next iteration of the rule.
PEG_FILENAME_PREFIX => SAVED,
UNPEG_FILE_TYPE: This is the type of pegged files that are being handled. This may be set to either: HEADER, where the datetime and duration fields are contained in the header, and apply to the whole file. DATA, where the datetime and duration fields are contained in the pif data rows, and apply individually to the pif data row.
KEEP_RAW_GRANULARITY: This option decides what happens when a PIF file is missing. A PIF file is determined as missing when the previous PIF end time does not match the current PIF start time. When this occurs a decision must be made on whether to output a LIF, with the values calculated with the available files. If KEEP_RAW_GRANULARITY is set to TRUE, then a LIF file will be created, ignoring the missing PIF. If KEEP_RAW_GRANULARITY is set to 0, then no output LIF will be produced to preserve the expected raw performance data granularity.
KEEP_RAW_GRANULARITY => TRUE,
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-65
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
CALCULATE_ROLLOVER
DEFAULT_NULL_VALUE
131
MAX_PEG_PIF_AGE: This formula is used to calculate the maximum time difference that is allowed between the previous timestamp and current timestamp, before the previous is considered to have expired. This may be the date and time counters in either the header, or counter data depending on the type of unpegging being done. This formula should include the DURATION tag, because the maximum age should be calculated as a factor of this. Duration is calculated as the difference between the start and end time of the current PIF. For example, if the start time of the file is 13:15, the end time is 13:30, the MAX_PEG_PIF_AGE will be 120 minutes (15*4+60). If the file has expired, the previous PIF will be deleted and the current saved for the next run.
MAX_PEG_PIF_AGE => DURATION*4+60,
CALCULATE_ROLLOVER: This option decides whether or not the unpegged value is calculated when rollover occurs. Pegged counter values always increase until they reach their maximum size (for example, 232-1), and are then rolled over to 0. If CALCULATE_ROLLOVER is set to TRUE, the UNPEGGER rule will calculate the rollover value, based on the maximum configured in PEG_COUNTERS.
4-66
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
If CALCULATE_ROLLOVER is set to 0, a NULL value will be inserted into the LIF for the counter value when rollover occurs.
CALCULATE_ROLLOVER => TRUE,
DEFAULT_NULL_VALUE: This is the value to use for the null value when required in the LIF output.
DEFAULT_NULL_VALUE => NULL
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-67
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
132
DATETIME_COUNTERS: This hash supports three options used to calculate the time difference between current and previous peg files, and the measurement duration. The date and time counters can be in almost any format as they are parsed using a DateTime module, which is extremely flexible in it accepted format. One of these combinations must exist in the PIF. Start date and start time and end date and end time. This hash maps the counters, STARTDATE, STARTTIME, ENDDATE, and ENDTIME to the counter names for the start and end date and times in the PIF header.
DATETIME_COUNTERS => { STARTDATE => 'StartDate', STARTTIME => 'StartTime', ENDDATE => 'EndDate', ENDTIME => 'EndTime', }
Start date and start time, and duration This hash maps the counters STARTDATE, STARTTIME, and DURATION to the counter names for the start date, start time and duration in the PIF header.
4-68
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
The optional FORMAT entry determines how the DURATION value is interpreted. Valid values are seconds and minutes. Defaults to minutes.
DATETIME_COUNTERS => { STARTDATE => StartDate, STARTTIME => StartTime, DURATION => gp, FORMAT => seconds, }
End date and end time, and duration This hash maps the counters ENDDATE, ENDTIME, and DURATION to the counter names for the end date, end time and duration in the PIF header.
DATETIME_COUNTERS => { ENDDATE => EndDate, ENDTIME => EndTime, DURATION => gp, FORMAT => seconds, }
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-69
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
PEG_COUNTERS
133
COUNTERS_TO_SORT_ON: This is the list of counters in the data rows used to map the row in the previous PIF to the current, when calculating the unpegged values:
COUNTERS_TO_SORT_ON => [ qw (GgsnFunction GiIsp) ],
PEG_COUNTERS: This is a hash containing a list of REs. Any counters in the input data, which match the RE, will be unpegged. The hash value specifies the rollover value for the counter. For 32-bit counters this should be set to 4294967295. For 64-bit counters this should be set to 18446744073709551615.
4-70
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
REDUNDANT_DATA_COUNTERS
134
The optional configuration entries for the UNPEGGER rule are: REDUNDANT_HEADER_COUNTERS: This is a list of counters omitted from the header in the output data. REDUNDANT_DATA_COUNTERS: This is a list of counters omitted from the data rows of the output data.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-71
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
135
See the sample configuration used in the sample application of the rule (see page 4-59).
4-72
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Student Exercise
IBM Software Group | Tivoli software
Student Exercise
Following the guidance of your instructor, check the UserConfig.pm file and save a copy in your personal folder.
Using the standard rules described in this unit, customize your UserConfig.pm file to have the best possible behavior of your Gateway. IBM Software Group | Lotus software
Discuss the results in class. Your instructor will use the best UserConfig.pm file in the actual Framework installation.
136
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
4-73
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Summary
IBM Software Group | Tivoli software
Summary
You should now be able to:
Identify the standard Post-Parser rules. Explain the following rules:
ACCUMULATE AGGREGATE_LINE BATCHFILES CVAL_MANIP DATALINE_WHERE FILE_SPLIT INFOINSERT JOIN MERGE_RECORDS PERLIZE PIF_2_OUTPUT PIF_REMOVE SPLIT_RECORDS UNPEGGER
137
4-74
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
A-1
139
The amount, size, and complexity of data to be processed varies from vendor to vendor. Therefore, performance of different vendor Gateways can vary widely. Any released vendor Gateways are profiled for performance problems as part of the testing process. However, there can be issues with server installations and configurations that require performance tuning. This appendix details some common changes that can improve the performance of a Gateway. Before applying any changes, ensure that the Gateway is putting out the correct LIF data for loading. It is also advisable to make a reference copy of the output, so that the resulting data from any changes can be validated. Minimize the number of intermediate and output files: Typically, a Gateway can process fewer larger files with more data more efficiently than a greater number of smaller files with a smaller number of records. For example, it is preferable to process a set of TRX counters in one large file per BSC, rather than a number of smaller files per CELL, even if it means that hierarchy data is duplicated. This is less of an issue if using PIF data is cached in memory. It may also have an impact on the loadmap configuration. Use the BATCHFILES rule in preference to the PIF_2_OUTPUT rule: The BATCHFILES rule can be used to create larger output PIF and LIF files, where multiple files matching the same pattern are joined into one larger file. This should be used in preference to the PIF_2_OUTPUT rule, because it produces less files, reducing I/O. It is also more efficient for subsequent loading.
Copyright IBM Corp. 2007
A-2
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
Use tmpfs file systems, or cache PIF data in memory. Use parallel processing. IBM Software
140
If putting out to multiple directories, use the Transfer Stage: For parallel PM/ NPR installations it is often required to output the performance data to two destinations. Rather than using multiple instances of the same rule to output to two directories, use the Transfer Engine or external scripts to distribute the files. Use tmpfs filesystems or cache PIF data in memory: Performance data goes through a number of transformations before final output. Disk I/O can be a major bottleneck. Use either tmpfs file systems or configure the PIF data to be cached in memory. This can offer significant savings. Use parallel processing: If the server has more than one processor, configure parallel processing for rules which meet the guidelines detailed. This can also be beneficial if the disk is slow, as multiple reads and writes can be queued to the disk. Use the PERLIZE rule for complex operations: If there is a specific counter manipulation or calculation requirement which requires a number of transformations, use the PERLIZE rule to configure it in a single function, rather than write specific rules. End the Gateway process gracefully: If required, the execution chain can be stopped gently by creating an empty file named stop_gateway in the input directory. Gateway will stop the current engine stage (does not finish parsing all remaining raw files) and proceed to post parsing stage. The remaining raw files will be parsed when the Gateway is restarted.
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
A-3
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.
A-4
IBM Tivoli Netcool Performance Manager for Wireless 9.1: Gateway Fundamentals
Course materials may not be reproduced in whole or in part without the prior written permission of IBM.