Softpanorama
(slightly skeptical) Open Source Software Educational Society

May the source be with you, but remember the KISS principle ;-)

Google   

wtdbclear

Usage: wtdbclear [-e] [-l] [-f] [-w] [-c <class>] [-r <severity>] [-s <status>] [-t <seconds>] [-p FALSE] [-a <# rows to commit after>]

Note: Never use if TEC server is running.

wtdbclear -left 86400 #  clears out all the entries in TEC older then 24 hours.

wtdbclear -let  86400 #  clears out all events older then 24 hours if everything is OK.  Never use if TEC server is running.

Good troubleshooting steps:

Some sample numbers.

Old News ;-)

IBM - Reading the wtdbspace at TEC 3.8 and 3.9

Question

Since the database schema has changed at version 3.8 how do you tell quickly that your TEC database is in a healthy state?

Answer

Firstly you can run wtdbspace with the -T option to limit the output to the TEC related data and the TEMP areas.

If any tablespace other than the TS_TEST_DATA is getting to about 80% you would want to run wtdbclear.

The new third section of wtdbspace shows you what tables are on which table spaces, and knowing which tables are cleared with wtdbclear and its options, you can decide which options of wtdbclear to use.

For example:
If only the TS_REC_LOG is getting full you can just run wtdbclear -lt 0.
If the TS_EVT_REP or TS_SLOTS_TASK are filling up you would need to do wtdbclear -et 0.
If TS_LONG or TS_INDEXES are getting full you'd probably need to run wtdbclear -elt 0 as all the TEC tables use those tablespaces.

Be aware that the table space names are customizable and so may not be TS_REC_LOG, etc.

IBM IBM Tivoli Enterprise Console V3.9 Implementation

Why is the wtdbclear binary preferred over the wtdbclear.pl Perl script?

  1. The wtdbclear binary accesses the new database stored procedures to improve rate of database clearing.

IBM - Running wtdbclear.pl, events queue up in TEC

tips on improving the performance of wtdbclear.pl -- sometimes the wtdbclear.pl processes run for hours as well as the queuing of events  

Cause need to use options on wtdbclear.pl  

Solution To improve the running of wtdbclear.pl processes, perform the following steps:

1. Set the -b parameter higher than the number of events to be retrieved.

NOTE: This makes the database read happen only once and not contend with the deletes. If this is not done, the program reads '-b buffer' amount of events, starts the deletes, reads the next set of events, and so on. Either process can get locked out at this point because of RIM internal hangs.

TEC 3.6.1-TMF-0049 fixes this problem by having separate 'agent' processes servicing each program. This is no longer an issue after patch 3.6.1-TEC-0006 is applied, as this patch changes wrimsql to read in all requested records in one pass.

2. Set the -a option low enough to get more frequent commits.

NOTE: The commit will free all held locks. This helps avoid contention and lock escalation to table or database locks.

3. Run clears for the reception log in a different job than the event repository clears.

NOTE: This helps avoid locks being set by one user on both tables at once and ending up with lock escalation to the database level. It would be best to only have one wtdbclear.pl job running at any given time.

4. Ensure the database's locking parameters are tuned well.

NOTE: Properly tuned databases are essential for applications with updating.

 

Reference

Clears events from the event database.

Syntax

wtdbclear [-e] [-f] [-l] [-c classname] [-r severity] [-s status] -t seconds [-a records] [-p TRUE | FALSE] [-w]

Description

The wtdbclear command deletes events from the reception log, event repository or both. Which tables it deletes rows from depends on the options specified.

The wtdbclear command must be run from the event server.

Authorization

senior

Options

-a records
A commit is performed after x number of records are deleted.
-c classname
Clears events with the specified class name.
Note:
You cannot use the this option when clearing the reception log with the -l option.
-e
Deletes events in the event repository, task repository, and extended event attribute table. You must use the -t option to specify the age of the events to be deleted.
-f
Forces deletion of QUEUED and WAITING events from the reception log.
Note:
Never use this option while the Tivoli Enterprise Console server is running.
-l
Deletes events in the reception log. You must use the -t option to specify the age of the events to be deleted.
Note:
You cannot use the -c, -r, or -s options when clearing the reception log with this option.
-p TRUE | FALSE
Indicates whether the stored procedure, tec_p_clear_events, should be called. The default value is TRUE. The tec_p_clear_events procedure improves the runtime performance of wtdbclear.
-r severity
Clears events with the specified severity. The valid default values are FATAL, CRITICAL, MINOR, WARNING, HARMLESS, and UNKNOWN.
Note:
You cannot use the this option when clearing the reception log with the -l option.
-s status
Clears events with the specified status. The valid default values are OPEN, CLOSED, RESPONSE, and ACK.
Note:
You cannot use the this option when clearing the reception log with the -l option.
-t seconds
Deletes events that are older than the specified number of seconds. This option is required.
-w
Forces deletion of events that have not yet been processed by the data warehouse extract, transform, and load (ETL) process. If the Tivoli Enterprise Console warehouse enablement pack is installed, events in the repository are periodically processed by the warehouse ETL process. If you issue the wtdbclear command and specify a time window that includes events that have not yet been processed by the warehouse ETL process, the wtdbclear command returns a warning message. Use the -w option to force deletion of these events.

Notes

If more than one of the -c, -s, and -r options is specified, the result is a logical OR.

Examples

The following example deletes all events from the event repository: 
wtdbclear -e -t 0

The following example clears the entire database: 

wtdbclear -elf -t 0

The following example disables the use of the stored procedure: 

wtdbclear -elt 0 -p FALSE

The following example, because it specifies more than one of the -c, -s,, and -r options, deletes all events that have a severity of HARMLESS or all events that have a class value of TEC_Start:

wtdbclear -r HARMLESS -c TEC_Start -et 0

See Also

wdbmaint.sh, wtdumper, wtdumprl, wtdumptr

 

Copyright © 1996-2008 by Dr. Nikolai Bezroukov. www.softpanorama.org was created as a service to the UN Sustainable Development Networking Programme (SDNP) in the author free time. Submit comments This document is an industrial compilation designed and created exclusively for educational use and is placed under the copyright of the Open Content License(OPL). Original materials copyright belong to respective owners. Quotes are made for educational purposes only in compliance with the fair use doctrine.

Standard disclaimer: The statements, views and opinions presented on this web page are those of the author and are not endorsed by, nor do they necessarily reflect, the opinions of the author present and former employers, SDNP or any other organization the author may be associated with. We do not warrant the correctness of the information provided or its fitness for any purpose.

Last modified: March 15, 2008