Tivoli 技术专家沙龙活动

Tsm Sql Interface

周硕基
[email protected]
 IBM Software Group | Tivoli software

Overview

•Tsm sql interface 介绍

•Tsm sql 在问题诊断中的案例分析
•Tsm sql 在做信息统计时常用的一些语句举例
•Tsm sql 其他

2
 IBM Software Group | Tivoli software

Tsm Sql Interface 介绍

 Supports the SQL SELECT query only

 Requires a “minimum” of 4MB of free space in the database(Some

conditions, such as the ORDER BY clause, the GROUP BY clause, and
the DISTINCT operator, may require additional temporary table storage.)

 Complicated queries may take a long time to complete and can interfere
with server operations(such as select * from backups)

 You cannot issue SELECT queries from the server console (admin
command line only )

3
 IBM Software Group | Tivoli software

Tsm Sql Interface 介绍

 “Mostly” conforms to standard SQL
 Subset of the SQL92 and SQL93 ANSI standards

 Does NOT support:

 UNION ( 合集）
 INTERSECT （交集 , 左右两边的两个查询都返回的所有非重复值 ）
 EXCEPT （从左查询中返回右查询没有找到的所有非重复值）
 Correlated subqueries ( 相关子查询 )
 Semicolon cannot be used as a command terminator

4
 IBM Software Group | Tivoli software

SELECT 语法
 SELECT column|expression [,n..]
FROM tablename {,n…}
 column refers to a column in a table
 * is allowed as a wildcard to select all columns in a table
 expression refers to functions that allow you manipulate the data being
returned
 [,n…] indicates that you may specify one or more columns or
expressions
 FROM clause indicates which table to search
 You can specify one or more tablenames

5
 IBM Software Group | Tivoli software

Tsm Database Catalog

 Tsm has three system catalog tables so that you can view the tables,
columns and enumerated data types available
 SYSCAT.TABLES

 SYSCAT.COLUMNS( 可以查询到表的 index 列 , 在检索时使用 index 列 , 可

以加快检索速度 )
select colname,index_keyseq from columns where
tabname='VOLUMES'

 SYSCAT.ENUMTYPES

6
 IBM Software Group | Tivoli software

Viewing the Tsm System Catalog

 To view available table names:
SELECT * FROM SYSCAT.TABLES

 To view all column names within tables

SELECT tabname,colname FROM SYSCAT.COLUMNS
 To view the valid values and order for enumerated types
SELECT * FROM SYSCAT.ENUMTYPES

7
 IBM Software Group | Tivoli software

Manipulating the Results

 You may not want all of the data in all of the columns all of the time so:

 SQL provides clauses, operators, and functions

 These allow you to sort, order, filter, and compute the data on a select
command

8
 IBM Software Group | Tivoli software

Clauses, Operators, and Functions

 ALL CURRENT_USER MAX
 ANY DISTINCT MIN
 AVG EXISTS MOD
 AS EXTRACT NULL
 BETWEEN FROM ORDER BY
 CASE GROUP BY POSITION
 CAST HAVING SOME
 COUNT IN SUBSTRING
 CURRENT_DATE JOIN SUM
 CURRENT_TIME LIKE TRIM
 CURRENT_TIMESTAMP WHERE

*Supported by TSM SELECT

9
 IBM Software Group | Tivoli software

WHERE Clause
 The WHERE clause allows you to filter out rows from the results

 I want this, this and this, but only where this condition is true

 I want to see all the volumes on which the client called test1 has data:

SELECT NODE_NAME,VOLUME_NAME FROM VOLUMEUSAGE

WHERE NODE_NAME=‘TEST1’

10
 IBM Software Group | Tivoli software

Operators (use with WHERE)

 Comparison Operators
is 、 is not 、 = 、 <> 、 < 、 > 、 <= 、 >=
 Logical Operators
Logical operators separate two or more conditions in the WHERE
clause
LIKE is used with the wildcard % to match all occurrences
SELECT * FROM NODES WHERE NODE_NAME LIKE ‘%T%’

AND means that the expressions on both sides must be true to return
TRUE
SELECT * FROM NODES WHERE NODE_NAME=‘TEST1’ AND
PLATFORM_NAME=‘WinNT’

11
 IBM Software Group | Tivoli software

More Logical Operators

 You can use OR to sum up a series of conditions. If any of the
comparisons is true, OR returns TRUE
SELECT * FROM NODES WHERE NODE_NAME=‘TEST1’ OR
‘TEST2’
 Use IN to replace multiple OR’s
SELECT * FROM NODES WHERE
NODE_NAMEIN(‘CARROL’,’DODSON’,’LEWIS’,’CHARLES’)

 Use BETWEEN to get a range

SELECT NODE_NAME FROM OCCUPANCY WHERE LOGICAL_MB
BETWEEN 5000 AND 10000

12
 IBM Software Group | Tivoli software

ORDER BY Clause
 The ORDER BY clause is used to sort the rows priorto displaying them:
SELECT NODE_NAME, PLATFORM_NAME FROM NODES
ORDER BY PLATFORM_NAME

 You can specify that the results be sorted in ascending or descending

order:
SELECT NODE_NAME, TYPE, FILESPACE_NAME,LOGICAL_MB
FROM OCCUPANCY ORDER BY LOGICAL_MB DESC

13
 IBM Software Group | Tivoli software

Functions
 Functions allow you to aggregate data and operate on strings, numeric
and date and time values

 Aggregate functions perform operations on values from selected rows to

produce a singlevalue.
They include COUNT(*), SUM, AVG, MAX, and MIN.

COUNT(*) is useful for finding the number of rows that match a query.

select count(*) from libvolumes

14
 IBM Software Group | Tivoli software

Timestamp and CAST Functions

 Example of date/time and cast function –displays nodes that have not
accessed the server in a specified ($1) number of days
 CAST’s the timestamp as decimal forprocessing
SELECT NODE_NAME,LASTACC_TIME FROM NODES WHERE
CAST((CURRENT_TIMESTAMP-LASTACC_TIME)DAYS as decimal)
>= 0
(This sql’s sample output will be:
NODE_NAME LASTACC_TIME
------------------ ------------------
TEST1 2007-08-19
22:06:27.000000
)

15
 IBM Software Group | Tivoli software

GROUP BY Clause
 The GROUP BY clause allows you to combine the rows being selected
into logical groups
 Normally used with aggregate functions
SELECT NODE_NAME, SUM(NUM_FILES) AS #_OF_FILES,
SUM(LOGICAL_MB) AS TOTAL_MB FROM OCCUPANCY GROUP
BY NODE_NAME

(this sql’s sample output will be:

NODE_NAME #_OF_FILES TOTAL_MB
------------------ ----------- ---------------------------------
TEST1 15 4.32
)
 When using aggregate functions, you need to
name the columns (i.e. AS #_OF_FILES)

16
 IBM Software Group | Tivoli software

HAVING Clause
 HAVING always follows the GROUP BY clause

 Use the HAVING clause to filter the results of the GROUP BY clause
SELECT NODE_NAME, SUM(NUM_FILES) AS #_OF_FILES,
SUM(LOGICAL_MB) AS TOTAL_MB FROM OCCUPANCY GROUP BY
NODE_NAME HAVING SUM(LOGICAL_MB)>1

17
 IBM Software Group | Tivoli software

Joining Tables
 Helps you see how data relates between tables

 There are different types of joins, depending on the data you are trying
to relate

 Use an alias to specify which column you want to display when joining
tables with columns of the same name

 The keyword DISTINCT specifies only unique rows will be retrieved and
prevent duplicates

18
 IBM Software Group | Tivoli software

JOIN Example
 To see which schedules a node is associated with join ASSOCIATIONS
with CLIENT_SCHEDULES
 DOMAIN_NAME and SCHEDULE_NAME are common columns

 We’ll use the alias C.SCHEDULE_NAME, to indicate which

SCHEDULE_NAME to return

select tabname,colname from columns where tabname in

('ASSOCIATIONS','CLIENT_SCHEDULES') and colname in('DOM
AIN_NAME','SCHEDULE_NAME')
the output will be:
TABNAME COLNAME
------------------ ------------------
ASSOCIATIONS DOMAIN_NAME
ASSOCIATIONS SCHEDULE_NAME
CLIENT_SCHEDULES DOMAIN_NAME
CLIENT_SCHEDULES SCHEDULE_NAME

19
 IBM Software Group | Tivoli software

Join Example continued

 SELECT DISTINCT NODE_NAME,C.SCHEDULE_NAME FROM
ASSOCIATIONS A, CLIENT_SCHEDULES C WHERE
A.SCHEDULE_NAME=C.SCHEDULE_NAME

 This is an example of an inner join or equi-join

 The goal is to match the values of a column in one table to the

corresponding values in the second table (schedule_name)

20
 IBM Software Group | Tivoli software

Tsm sql 案例一

 用户在用客户端恢复文件的时候，报告此客户端上 2006 年 12 月 31 日归
档的 /tmp/test1 文件恢复失败
 查询 actlog 表查看具体报错信息 (or q act begind=xxx sear=err/fail)
 定位 /tmp/test1 文件
有几个归档：
select node_name,filespace_name,filespace_id,archive_date, class_name
from archives where ll_name='test1‘
此归档文件存放在哪些卷：
select volume_name,node_name,type,filespace_name from contents where
type='Arch' and file_name like '%test1'

21
 IBM Software Group | Tivoli software

Tsm sql 案例一 continued

 通过刚才的两个 sql 查询，我们可以定位需要恢复的文件一共做过几次归
档，各个归档存放在哪一个卷上
 思考一：我们如何通过一个查询连接这两个表进行恢复文件定位？
（ substr(contents.file_name) 和 archives.ll_name)
 下一步如何处理？
 寻找其他时间点的归档进行恢复
 对 volume 状态重置 (q vol xx f=d or select * from volumes where access like
xx)
 做 audit libr xxx or audit vol xxx
…

 思考 : 恢复不了的文件是做的备份而不是归档还要注意哪些地方？

select state,backup_date ， class_name from backups where xxx

注意活动非活动版本
注意不同的管理类
…
22
 IBM Software Group | Tivoli software

Tsm sql 案例二

 用户在做 backup db 做到一半，由于某种原因中断，怎么样让此 db 备份
磁带变回 scratch 状态？
select volume_name as "DBB Full Tapes", date_time as "Time
Created", current_timestamp as "Current D&T" from volhistory where
cast ((current_timestamp-date_time)days as decimal) >=0 and
type='BACKUPFULL' order by date_time

DBB Full Tapes Time Created Current D&T

------------------ ------------------ ------------------
abc123 2007-08-20 2007-08-20
19:49:49.000000 19:51:28.000000
abc321 2007-08-20 2007-08-20
19:49:49.000000 19:51:28.000000

next?
 Del volh t=dbb

23
 IBM Software Group | Tivoli software

Tsm sql 案例三

 用户反映在一个单节点归档 , 两驱动器的带库备份系统中 , 每天备份量大
小基本相同 , 虽然归档时间设置成 30 天 , 但存储池中磁带空间不见循环
使用 ?
一般判断 , 应该是存储部分的故障
select date_time,msgno,message from actlog where severity=‘E‘ or
severity=’W’ 检查 act log 相应的错误信息
select drive_name,online from drives
select destination_name,destination_type,library_name,device,online
from paths
常见故障分析 :one drive offline,one path offline,path 定义错误…

24
 IBM Software Group | Tivoli software

Tsm Sql 常用统计语句

 在我们抓取用户信息做服务方案 , 做带库扩容方案 , 做时间估算…
 有多少磁带逻辑状态不正常 ?
select VOLUME_NAME,ACCESS from volumes where access
='UNAVAILABLE'
 带库里目前有多少 scratch 磁带 ?
select count(*) as Scratch_count from libvolumes where status='Scratch'
 在改变存储池回收回收阀值时有多少个磁带会做回收处理？
select count(*) from volumes where stgpool_name='poolname'and
upper(status)='FULL' and pct_utilized <%%
 各存储池卷分配情况
select stgpool_name,devclass_name,count(*) as "VOLUMES" from
volumes group by stgpool_name,devclass_name

25
 IBM Software Group | Tivoli software

Tsm Sql 常用统计语句二

 按照节点统计文件和文件大小
select node_name, sum(logical_mb) as Data_In_MB, sum(num_files) as
Num_of_files from occupancy group by node_name order by
node_name
 过去 24 小时 tsm 数据备份总量
SELECT entity AS "Node name",CAST(sum(bytes/1024/1024) AS
char(29)),1,10) AS "start date", SUBSTR (CAST(min (start_time) AS
char(29)),12,8) AS "start time", SUBSTR (CAST(max (end_time) AS
char(29)),1,10) AS "end date", SUBSTR (CAST(max (end_time) AS
char(29)),12,8) AS "end time" FROM summary WHERE
activity='BACKUP' AND start_time>=current_timestamp - 24 hours
GROUP BY entity

(activity 可取值 :'TAPE MOUNT',

'FULL_DBBACKUP','INCR_DBBACKUP',
'EXPIRATION', 'RECLAMATION', 'MIGRATION', 'MOVE DATA',
'STGPOOL BACKUP', 'BACKUP', 'RESTORE', 'ARCHIVE' and
'RETRIEVE'. )

26
 IBM Software Group | Tivoli software

Tsm Sql 常用统计语句三

 列出使用自动调度的客户端节点
SELECT node_name FROM nodes WHERE node_name IN (SELECT
node_name FROM associations) （ 子查询非相关子查询）

other:
events?
策略相关 ?

27
 IBM Software Group | Tivoli software

Tsm sql other

 Odbc?
 scripts(scripts.smp)

28
 IBM Software Group | Tivoli software

Tsm Sql 参考
 Help select

 http://www.redbooks.ibm.com/abstracts/tips0010.html?Open

 http://thobias.org/tsm/sql/index.html

 http://www.lascon.co.uk/d005104.htm

29
 IBM Software Group | Tivoli software

Questions????

30