DESTA - Database files growing out of control

[RedLimey]

It looks like something is corrupt within our HPSIM/WEBES/SEA installation. Within the svctools\specific\desta\database\data\base folder there are folders named numberically. In particular folder, I have 3.5Gb of files and it's consuming any free space we have on that partition. We recently moved the page file off this partition (3+ Gb), so the rate of growth is abnormal. Question is why?

Any clues/pointers on how to troubleshoot this is greatly appreciated.

Thanks,
SR

[Teem]

Just upgraded to WEBES 6.0 last week and I'm seeing the exact same problem, except my folder is already up to 6 Gb and growing.

[RedLimey]

Before I closed down for the night, I stopped the DESTA service. This morning when I came in, I still showed low space on the partition (no surprises there). After starting the DESTA service with the intent of seeing what I could find in the form of logs and events, I noticed in fairly quick time that the DB files has cleared out. I now have my space reclaimed.

We actually have two instances of HPSIM/WEBES (the other managed by a different team) and coincidentally, BOTH befell the same symptom over the same period. Very interesting.

[Teem]

Red,

Was that the first time that you had bounced the service since install? I wonder if thats a one time logging setting that gets cleared by stopping and starting the service?

[RedLimey]

No - I recycled the service last night with no immediate indicator that DESTA was actually doing anything (I wasn't exactly patient to see if time would favor the recycle). Space never hit 0. I did see free space improve a little (say, 150mb to 370mb) before being re-consumed but did not ID what that was due to (desperate OS measure, like file compression?). Not finding any clues as to what was going on or what action to take, I decided to stop the service and hope the system didn't crash.

[Manfred Schilling]

I think it's the same problem here. In the last weeks Postgres-DB is growing out of control in table "pg_largeobject" of "destadb" (actual 45 Mio records, 13 GB size). We installed WEBES 5.6 Update 2, but nothing changed. Any idea ?

[RedLimey]

No answers here. We did have a repeat last night and this morning the partition was down to 10mb. I've killed DESTA and deleted the files for now. We're trying to get some feedback from an internal source at HP. If we find something, I'll pass it along. In the meantime, if someone else knows why/has the cure to whatever is screwing up with Webes, please pass it along.

Thanks,
-SR

[Manfred Schilling]

After this weekend additional 2,5 GB used in table pg_largeobject (16 GB, 55 Mio rec). We found, binary data are form EVA eventlogs. We decided to drop and recreate pg_largeobject, but was rejected by permission to drop system table. Any idea ?

[Teem]

HP recommendations doing it tonight will let you guys know how it goes.

1) Go to Start/Allprograms/HewlettPackard/Remote Support Software Manager/Remote Support Software Manager Interface
On the Packages list select WEBES 6.0 and then click on remove
2) When WEBES is removed go to and download the WEBES Cleanup Util and run it
3) Then go to Add Remove Programs and remove Postgresl and delete the remaining folder under the C://ProgramFiles/Postgresql.
4) When all that is remove just go back to the Remote Support Software Manager Interface and select install to WEBES 6.0, this will install the new Postgresql.
5) The server needs to be rebooted.

[Manfred Schilling]

HP told uns how to "reset" Database:

stop DESTA-service

Start ... -->PostSql 8.3 --> SQL Shell (psql)
Server [localhost]:
Database [postgres]:
Port [7950]:
Username [postgres]:
...
postgres=# drop database destadb;
DROP DATABASE
postgres=# drop database wbemdb;
DROP DATABASE
postgres=# drop database ed;
DROP DATABASE
postgres=# create database destadb with owner="WEBES" encoding='UTF8';
CREATE DATABASE
postgres=# create database wbemdb with owner="WEBES" encoding='UTF8';
CREATE DATABASE
postgres=# create database ed with owner="WEBES" encoding='UTF8';
CREATE DATABASE

start DESTA-service

Database grows overnight to > 1GB :-(

[RedLimey]

ZIM --- Useful info to have if the DB does ever need housekeeping, however as I think you're saying, with DESTA behaving the way it is, it all just comes back in quick fashion.

Teem --- Did a reinstall help any? That's still a poor solution from HP. They have no idea why this has started to happen with their product at multiple locations??

[Manfred Schilling]

Database still growing... Source is WEBES on CV-Server after an EVA8400 update to Version 09534000 (seem to be always the same entrys). How to reset EVA Controller-Eventlog ?

[Manfred Schilling]

Our problem is now solved by resetting cache for that EVA on the CV EVA-Server.

[RedLimey]

It turns out we too have an EVA that had a faulty component and was flooding alerts. Both our HPSIM instances were referencing it, hence both were symptomatic.

The following steps were supposedly used by an HP tech to reset the the DB for 6.0.

1. On the CMS launch SEA UI and login with "admin-adv"
2. Go to the Managed Entities Page and click the "Export" button. This will launch another window that displays the process of the Managed Entity data being saved in XML format.
3. Close the window when this completes
4. Close the SEA UI

Then from the CLI on the CMS:

5. stop the desta_service.

net stop desta_service

6. cd C:\Program Files\HP\svctools\common\database\bin
7. Run the following command:

psql -h localhost -U WEBES -d postgres -p 7950

8. The psql window will appear with the following command prompt:

postgres-#

9. To drop existing databases perform the following:

DROP DATABASE IF EXISTS destadb;

10. To recreate the databases perform the following:

CREATE DATABASE destadb WITH OWNER = "WEBES" ENCODING = 'UTF8';

11. To exit type: \q

12. Start Desta Service

net start desta_service

[Teem]

Didn't get a chance to use the HP recommendations. Server was rebooted after patching and the problem hasn't come back yet...still monitoring.

[datacenter]

Hi,

i tried to delete the table, because my desta db growing up 6GB, i stop the desta service but when i try to logon in DB i received the request for a password of WEBES user, i didn't know it, so i tried to logon with administrator of local machine, but nothing to do. is there a default username and password?

Thanks to all

[MikeWachholz]

So I got to step 7, and it's asking for a password? Does anyone know what the default WEBES password is???

Thanks!
Mike

7. Run the following command:

psql -h localhost -U WEBES -d postgres -p 7950