It's all about the answers!

Ask a question

HELP!!! How to clean-up DB on MS SQL 2005 EXPRESS


David Csikkel (1313516) | asked Mar 16 '10, 4:55 p.m.
Hello,

according to solving of
http://jazz.net/forums/viewtopic.php?t=9631 (still unresolved :-( )
we tried to switch logging from WARN to DEBUG mode and
probably it was reason why DB has started to grow and achieved limit size 4GB. We have found out that it was caused by two tables (in both more then 800.000 rows)

REPOSITORY.CHANGE_EVENT
REPOSITORY.ITEM_STATES

So we have switched logging back to WARN mode and
we have deleted all rows
which had EVENT_CATEGORY equal to 'SystemLog' from first table
and which had ITEM_TYPE_DBID equal to 28 from second one

Could it affect somehow Jazz?
Is DEBUG log mode the right reason of DB growing?
Does there exist another option how to decrease DB size?

Thank you in advance

David Csikkel
Consultant, Tieto

5 answers



permanent link
Anthony Kesterton (7.5k9180136) | answered Mar 17 '10, 5:47 a.m.
JAZZ DEVELOPER
Hello,

according to solving of
http://jazz.net/forums/viewtopic.php?t=9631 (still unresolved :-( )
we tried to switch logging from WARN to DEBUG mode and
probably it was reason why DB has started to grow and achieved limit size 4GB. We have found out that it was caused by two tables (in both more then 800.000 rows)

REPOSITORY.CHANGE_EVENT
REPOSITORY.ITEM_STATES

So we have switched logging back to WARN mode and
we have deleted all rows
which had EVENT_CATEGORY equal to 'SystemLog' from first table
and which had ITEM_TYPE_DBID equal to 28 from second one

Could it affect somehow Jazz?
Is DEBUG log mode the right reason of DB growing?
Does there exist another option how to decrease DB size?

Thank you in advance

David Csikkel
Consultant, Tieto


Hi David

I am not sure if removing the entries is good/bad but change_events and item_states sound like things you should NOT be removing. Perhaps someone from the repository team can comment?

RTC does a lot of compression internally - so lots of rows in not necessarily occupying a lot of space (I could be completely wrong about these specific entries in the db). Can you run a repository report (it should be one of the Reports in any of your projects) - this should tell you what is using up the space and what the compressed size is.

Are you able to move to another database, or get a version of SQL, that does not have the 4GB limit? If you fix the problem today, I suspect you will run out of space at some point - and your problem might be more complicated (there is a lot of real data in the db rather than lots of error message

regards

anthony

permanent link
David Csikkel (1313516) | answered Mar 17 '10, 10:30 a.m.
Sorry :-O I missed to note one important thing we reached these 4GB in the short period (2-3weeks) in not often used Jazz server (5 iterations+plans, ~250 work items) without any file in repository and any builds. It points to saving of logging information in DB (according to DEBUG mode and 'System log' value in REPOSITORY.CHANGE_EVENT).

T.Y.I.A

David

permanent link
Anthony Kesterton (7.5k9180136) | answered Mar 18 '10, 3:00 p.m.
JAZZ DEVELOPER
Sorry :-O I missed to note one important thing we reached these 4GB in the short period (2-3weeks) in not often used Jazz server (5 iterations+plans, ~250 work items) without any file in repository and any builds. It points to saving of logging information in DB (according to DEBUG mode and 'System log' value in REPOSITORY.CHANGE_EVENT).

T.Y.I.A

David


Wow - that is amazing how quickly it filled up!

Let's see what the repo team says. To make sure of this - please raise a defect here on jazz.net.

anthony

permanent link
Gary Mullen-Schultz (28725536) | answered Aug 26 '10, 5:44 p.m.
Was a work item submitted about this issue? I have a customer encountering the same thing.

Thanks, Gary

permanent link
Anthony Kesterton (7.5k9180136) | answered Aug 27 '10, 10:56 a.m.
JAZZ DEVELOPER
Was a work item submitted about this issue? I have a customer encountering the same thing.

Thanks, Gary


Not sure if a defect was raised - but since this thread started, I saw an instance where there were error events being generated - very quickly and for a long time. This caused the DB to fill up. Moving to 2.0.0.2 iFix3 seemed to have solved the problem (originally on 2.0.0.2 - no iFix)

anthony

Your answer


Register or to post your answer.


Dashboards and work items are no longer publicly available, so some links may be invalid. We now provide similar information through other means. Learn more here.