Lab Guide

Isilon InsightIQ
Lab Guide

EMC Solutions Group

Abstract
This Lab Guide serves as a guide for a demonstration of Isilon InsightIQ.

June 2012
Copyright © 2012 EMC Corporation. All Rights Reserved.

EMC believes the information in this publication is accurate as of its

publication date. The information is subject to change without notice.

The information in this publication is provided “as is.” EMC Corporation makes
no representations or warranties of any kind with respect to the information in
this publication, and specifically disclaims implied warranties of
merchantability or fitness for a particular purpose.

Use, copying, and distribution of any EMC software described in this

publication requires an applicable software license.

For the most up-to-date listing of EMC product names, see EMC Corporation
Trademarks on EMC.com.

All trademarks used herein are the property of their respective owners.

Isilon InsightIQ 2
Table of contents
Introduction ..........................................................................................................................................4
Purpose ............................................................................................................................................4
Scope ...............................................................................................................................................4

Technology overview .............................................................................................................................5

Isilon InsightIQ .................................................................................................................................5

Lab 1 – Using InsightIQ .........................................................................................................................7

Overview ...........................................................................................................................................7
Lab ...................................................................................................................................................7
Logging in to Insight IQ .................................................................................................................7
The DASHBOARD ..........................................................................................................................7
Performance Reporting .................................................................................................................8
Exploring Network Statistics ...................................................................................................... 11
Network throughput rates for a specific client. .............................................................................12
Network throughput rates for all clients .......................................................................................13
Network Throughput Rates going in and out of the cluster ............................................................14
Network Throughput Rates per Node ............................................................................................14
Network Throughput Rates per Interface ......................................................................................14
Network Throughput Rats per Protocol .........................................................................................15
Permalink .................................................................................................................................. 15
Performance Reporting .............................................................................................................. 16
Viewing Disk activity....................................................................................................................17
Creating a Performance Report. ...................................................................................................18
Managing Performance Reporting ................................................................................................22
File System Reporting ................................................................................................................ 23
Advanced Filters ..........................................................................................................................26
Exploring Data Properties of a cluster. .........................................................................................28
Comparing Data Changes. ...........................................................................................................32

Isilon InsightIQ 3
Introduction
Purpose This lab guide serves as a guide to a demonstration of Isilon InsightIQ.

Scope This Lab Guide provides:

 An introduction to using InsightIQ to monitor and analyze Isilon cluster
performance by viewing flexible, customizable charts that display historical
performance data.
 Labs to analyze data usage, data properties and other performance data at the
file-system level.

Isilon InsightIQ 4
Technology overview
Isilon InsightIQ
InsightIQ enables you to monitor and analyze Isilon cluster activity through flexible,
customizable chart views in the InsightIQ web-based application; these charts
provide detailed information about cluster hardware, software, and file-system and
protocol operations.
InsightIQ transforms data into visual information that emphasizes any performance
outliers, enabling you to quickly and easily diagnose bottlenecks or optimize
workflows.
With InsightIQ you can:
 determine whether a cluster is performing correctly.
 compare changes in performance across multiple metrics, such as CPU usage,
network traffic, protocol operations, and client activity.
 correlate critical cluster events with performance changes.
 determine the effect of workflows, software, and systems on cluster
performance over time.
 view and compare properties of the data on the file system.

By transforming complicated data into relevant, visual information, you can:

 determine a cause or validate a theory. You can view specific charts in order to
identify or confirm the cause of a performance issue. For example, if end users
were reporting client connectivity problems during certain time periods, you
could view InsightIQ charts that enable you to determine whether the issue was
occurring at the cluster level, the node level, the client level, the disk level, or
the network level. InsightIQ enables you to easily correlate seemingly unrelated
data across both present and historical conditions.
 measure the effects of configuration changes. If you modify your cluster
environment, you can determine the effect of those changes by comparing
charts that represent past, baseline performance with charts that represent
performance since the changes were made. For example, if you added 20 new
clients, you could view before-and-after charts that would indicate whether the
added clients had an effect on a specific aspect of system performance, such
as network throughput or CPU usage.
 optimize application workflows or performance. Reviewing InsightIQ charts can
help you identify bottlenecks or inefficiencies in your systems or workflows. For
example, if you wanted to ensure that all clients can access the cluster quickly
and efficiently, you could view charts that indicate whether some client
connections are significantly faster or slower than others. You could then apply
breakouts and filters as needed to identify the root causes, which could be
related to specific files that the clients are working with or the operations that
the clients are performing.
 analyze detailed operations data. InsightIQ charts, breakouts, and filters can
provide specific information about general cluster operations. For example, if

Isilon InsightIQ 5
 you recently deployed an Isilon cluster, you might want to view detailed charts
that illustrate how the cluster and its individual components are performing.
 forecast future needs. Reviewing past performance trends can help you predict
future trends and needs. For example, if you deployed an Isilon cluster for a
data-archival project six months ago, you might want to estimate when the
cluster will reach its maximum storage capacity. You could view and analyze
charts that illustrate storage-capacity usage by day, week, or month in order to
determine approximately when the cluster will reach capacity if current trends
were to continue.

Isilon InsightIQ 6
Lab 1 – Using InsightIQ
Overview This Lab is a walkthrough of the basics of using InsightIQ to monitor and analyze
Isilon cluster performance.

Lab Logging in to Insight IQ

Log in to InsightIQ:
1. Use RDP to connect to the IsilonIQ Window Client. Use the highest resolution
as the web interface is large. (See Figure 1)

Figure 1. Connecting With RDP

2. Click the FireFox icon to open the web browser (See Figure 2).

Figure 2. Launching the InsightIQ web interface

3. Enter the following username and password and click Log in (See Figure 3)
 Username: administrator
 Password: Password

Figure 3. Logging into InsightIQ

This opens the InsightIQ.

The DASHBOARD
1. Highlight the following on the DASHBOARD .(See Figure 4)

Isilon InsightIQ 7
 Figure 4. Viewing the DASHBOARD

 1. Aggregate view of data from all clusters

 2. The graphs on the right are recent trends
 3. This handle can be used to drag the cluster data up and down
 4. With the Performance details link, users can jump straight to
performance details on this cluster
 5. With the File system details links, users can jump straight to File system
details on this cluster
 6. When there is a large - icon, it can be clicked to collapse this data
Performance Reporting
1. Click PERFORMANCE REPORTING to view reporting options. (See Figure 5)
2. In the Apply this Report list box, select Cluster Performance. (See Figure 5)

Isilon InsightIQ 8
 Figure 5. Selecting a Performance Report.

3. In the Apply to this Cluster list box, select qa-long3. (See Figure 5)

Figure 6. Apply a report to a cluster

4. Click the calendar icon and select April 20th. (See Figure 7)

Isilon InsightIQ 9
 Figure 7. Specifying a report date

5. This lab is using pre-loaded engineering cluster data so it’s best to use the
Date Range shown in Figure 8
6. Once the date has been specified click View Report. (See Figure 8)

Figure 8. Viewing Reports

7. Expand Event Summary (See Figure 9) and browse though the displayed data.
Highlight the following:
 1. Health and general statistics of the cluster
 2. Zoom level can be quickly change here. (Warning: do not display data
before Apr 20 2011 after this step, client filters will not work)
 3. Auto Refresh can be toggled. This allows near real-time data to be
explored when connect to a live cluster
 4. Many filter options exist by using the Create/manage data filters link
 5. The report can be scheduled and even emailed
 6. The report can be customized from the default data displayed and many
customer reports can be made.

Isilon InsightIQ 10
 Figure 9. Viewing a specific report.

Exploring Network Statistics

1. On the External Network Throughput Rate graph, hover over any part to
display data at that point in time. The date and times may be changed by
dragging the graph left or right. (See Figure 10)
Note: To zoom in double click on the graph. To zoom out press the shift key
on your keyboard and double click the mouse.

Figure 10. Viewing External Network Throughput Rate

2. Click Client to show the clients with the most network throughput. (See Figure
11)

Isilon InsightIQ 11
 3. Click and drag the graph to show all the heat maps change dynamically. Each
client displays the percentage of all traffic being used by that client. (See
Figure 11)

Figure 11. Viewing client data

Network throughput rates for a specific client.

1. Click 10.53.192.53 to view more statistics for this client. (See Figure 12)

Figure 12. Viewing statistics for a specific client

2. In the Breakout by: section, click Op Class. (See Figure 13).

Isilon InsightIQ 12
 3. Hover over the heat maps to display data for this specific client at that point
in time. The percentage of all traffic being used by the client is displayed next
to each statistic. (See Figure 13).

Figure 13. Viewing client traffic

4. Clear the client specific filter by clicking the Client: 10.53.192.53

Network throughput rates for all clients
1. Hover over the heatmaps to display data for all clients at that point in time.
The percentage of all traffic being used by the client is displayed next to each
statistic. (See Figure 14). Note: With this information you can deduce that
SSD drives for metadata would be very beneficial because namespace_reads
are metadata operations.

Figure 14. Viewing network traffic for all clients

Isilon InsightIQ 13
Network Throughput Rates going in and out of the cluster
1. In the Breakout by: section click Direction. Hover over the heatmaps to
display data for that client at that point in time. Next to each direction the
percentage of traffic, going in that direction, is displayed. (See Figure 15).

Figure 15. Viewing direction of network traffic for a specific client

Network Throughput Rates per Node

1. In the Breakout by: section click Node. Hover over the heatmaps to display
data for that node at that point in time. Next to each node the percentage of
all traffic, that node is providing, is displayed. (See Figure 16).

Figure 16. Viewing network traffic for a specific node

Network Throughput Rates per Interface

1. In the Breakout by: section click Interface. Hover over the heatmaps to
display data for that interface at that point in time. Next to each interface the
percentage of all traffic, that interface is providing, is displayed. (See Figure
17).

Isilon InsightIQ 14
 Figure 17. Viewing network traffic for a specific interface

Network Throughput Rats per Protocol

1. In the Breakout by: section click Protocol. Hover over the heatmaps to display
data for that protocol at that point in time. Next to each protocol the
percentage of all traffic being delivered via that protocol, is displayed. (See
Figure 18).

Figure 18. Viewing network traffic by protocol

Permalink
1. Click PERMALINK at the top of the page. This allows you to send the link to
other administrators who can then see the same view. (See Figure 19)

Isilon InsightIQ 15
 Figure 19. Permalink to send link to other administrators

a. Highlight the fact that this URL can be copied and sent to other
administrators. Click OK. (See Figure 20)

Figure 20. URL which can be copied

Performance Reporting
1. In the Apply this Report list box select Disk Performance and click View
Report. (See Figure 21)

Isilon InsightIQ 16
 Figure 21. Viewing Disk Performance reports

Viewing Disk activity

2. In the Breakout by: section click Disk. Hover over the heatmaps to display
data for that drive. (See Figure 22)

Figure 22. Viewing Disk Performance per drive

3. Click show more to get more detailed information. Highlight how this visual
(Figure 23) shows OneFS writing data to many drives and not limiting the disk
activity to a small set of drives.

Isilon InsightIQ 17
 Figure 23. Viewing detailed Disk Performance

Creating a Performance Report.

1. At the top of the page click Create a new Performance Report. (See Figure 24

Figure 24. Creating new Performance Reports

2. Under Standard Reports click Cluster Performance Report Template. (See

Figure 25)

Isilon InsightIQ 18
 Figure 25. Selecting report type

3. Provide the following information to build the report: (See Figure 26)
 Performance Report Name: My Report
 Select Live Performance Reporting
 Select Scheduled Performance Report
 In the Subject of Report list box select qa-long4
 Select Email this report….
 For Report Recipients type [email protected]
 In the External Network Throughput Rate select Client (See Figure 27)
You can also move sections up and down by clicking and holding the handles
and moving up or down. This shows how flexible the reports are. (See Figure
26)

Isilon InsightIQ 19
Figure 26. Specifying report details

Isilon InsightIQ 20
 Figure 27. Specifying data to be reported

4. At the end of the page click Add another performance module. (See Figure
28)

Figure 28. Adding extra performance modules

Isilon InsightIQ 21
 5. In the Select a Module for this Position list box select another report to add.
Click Save. (See Figure 29)

Figure 29. Saving reports

InsightIQ confirms your performance report has now been saved. (See Figure
30)

Figure 30. Confirming report saved

Managing Performance Reporting

6. Under PERFORMANCE REPORTING click Managing Performance Reporting.
Show the options available for the report you just created. (See Figure 31

Isilon InsightIQ 22
 Figure 31. Managing Performance Reporting

7. Click Scheduled Performance Reports. Show the options available for

scheduled reports. (See Figure 32)

Figure 32. Scheduling Performance Reports

File System Reporting

1. Click FILE SYSTEM REPORTING. In the Apply this report list box select Data
Usage. (See Figure 33)

Figure 33. Selecting File System Reporting

Isilon InsightIQ 23
2. In the Select a reporting data list box, select the Sun Apr 17 2011 16:07:01
GMT-0700 (PDT) report. Click View Report. (See Figure 34)

Figure 34. Selecting report dates

3. Highlight the following in this report : (See Figure 35)

 1. Health and general statistics of the cluster.
 2. Hover over any slice of the pie and you will see the directory.
 3. You can click any column heading to re-sort the data.

Figure 35. Viewing File System Reports

4. Make sure you sort the data by Logical Size (default sorting) before
continuing. (See Figure 36).

Isilon InsightIQ 24
 Figure 36. Viewing reports on /ifs/

5. Click on the biggest slice of the pie (/ifs/data) (See Figure 36).

Figure 37. Viewing reports on (/ifs/data/)

6. Click on the biggest slice of the pie (/ifs/data/5.0.x) (See Figure 38)
7. Click data/5.0.x/, this will filter all the data below to just that path. (See
Figure 38)

Figure 38. Filtering reports to data/5.0.x/

Isilon InsightIQ 25
Advanced Filters
1. Click Create/manage data filters. (See Figure 39 )

Figure 39. Managing data filters

2. Click Add Rule. (See Figure 40)

Figure 40. Adding rules

3. In the list box under Type highlight all the options available. (See Figure 41)
Using the Manage menu, filters can be saved and applied to any data in any
view.

Figure 41. Showing options available

4. Click Hide filter details to hide the filter option. (See Figure 42)

Isilon InsightIQ 26
 Figure 42. Hiding filters

5. To find the largest file in the /ifs/data/5.0.x directory hover over

…com.flat.vmdk in the Top 1000 Files by Largest Logical Size. (See Figure 43)

Figure 43. Viewing largest files in /ifs/data/5.0.x

6. In the Top 1000 Files by list box highlight other available options for finding
files. (See Figure 44)

Figure 44. Viewing other options for finding files

7. Scroll down the page to view where directories can also be filters. In the Top
1000 Directories by list box highlight the options available. (See Figure 45)

Isilon InsightIQ 27
 Figure 45. Viewing other directory options

8. Clear the path by clearing the filter at the top of the page. (See Figure 46)

Figure 46. Clearing path filters

Exploring Data Properties of a cluster.

1. In the Apply this report list box, select Data Properties.(See Figure 47)
2. In the Select a reporting date list box select Sun Apr 17 2011 16:07:01 GMT-
0700 (PDT). (See Figure 47)
3. Click View Report. (See Figure 47)

Figure 47. Viewing data properties of a cluster

4. Scroll down the page to view the data displayed. Show that the number of
files on the cluster between 8KB and 128KB is 524,398,436 by hovering the
mouse over the graph. (See Figure 48)

Isilon InsightIQ 28
 Figure 48. Viewing data displayed

5. In the Breakout by: section of File Count by Physical Size, click Modified
Time. Highlight that 295M of the files between 8KB and 128KB are over 2
years old. (See Figure 49)

Isilon InsightIQ 29
 Figure 49. Viewing data over various times

6. Click 2 years + to filter data to 2 years or older. (See Figure 50). Highlight how
all the graphs are now filtered to files that are 2 years or older.
7. Hover over the graph to highlight how the same data retrieved in step 5 can
be retrieved here.

Figure 50. Viewing data 2 years old

8. In the Breakout by: section, click Path. Hover over ...x/logserver/qa.You will
see the full path where the largest number of old files is located. (See Figure
51)

Isilon InsightIQ 30
 Figure 51. Viewing full paths

9. In the Breakout by: section, click Disk Pool. Highlight that most of the files
are in the NL36000x pool, which is composed of NL nodes. This is where they
should be stored since the NL nodes are more cost effective per TB and since
they are old files, they should be stored there. (See Figure 52)
10. Click ctime:2 years to remove the filter. (See Figure 52)

Figure 52. Removing timeline filters

11. In the Breakout by: section click Extension. Highlight that the majority of files
are svn-base files. This is correct because this is the cluster that stores source
code. (See Figure 53)

Isilon InsightIQ 31
 Figure 53. Viewing source code files

Comparing Data Changes.

1. In the Compare to list box, under Select Your File System Report, at the top of
the page, select Thu Sep 02 2010 16:14:00 GMT-0700 (PDT).Click View
Report. (See Figure 54)

Figure 54. Comparing reports

2. Show how all the same data is displayed and note all the differences from the
report run 6 months before. (See Figure 55)

Isilon InsightIQ 32
 Figure 55. Viewing differences between reports

3. In the Breakout by: section click Disk Pool. (See Figure 56) Highlight that
there is 362M more files on the NL36000x pool and 271M less files in the
default pool. From this information you can deduce that SmartPool Files Pool
Polices have been setup to move data to the cheaper NL nodes.

Figure 56. Viewing Disk Pools

That completes the demonstration of

Isilon InsightIQ

Thank You

Isilon InsightIQ 33