[Openvas-discuss] Openvas scans not completing and GSA very slow

Matthew Mundell matthew.mundell at greenbone.net
Fri Sep 16 09:58:01 CEST 2011


> The error logs I can find seem to indicate the scan abruptly stops after
> about five hours.  Due to the degraded GSA performance I'm wondering if this
> is a database issue somehow?  Is it possible the response time on the
> database is so bad that the scan is timing out or hitting a maximum run
> time?  I've run VACUUM and rebuild on the SQLITE database without any
> apparent change.  Our database is approximately 450MB at the moment.

The bigger databases I've seen are about a 100MB.

The slow part in the database access is usually applying the overrides to
the report results.  This has to be done for the main page to get the
counts and trends.  I guess this is what you are seeing with it getting
slower and slower.

Am a bit doubtful that the slow database would affect the overnight scan,
but maybe it does.  Do you have a duration on the schedule?  If this scan
takes too long this would stop it.

> Does this sound like two symptoms of the same problem or am I dealing with
> two separate issues?  Is there a way to prune old records from the sqllite
> database or perhaps some indices I can create to get it to respond faster?

The best way to reduce size is to remove some of the reports from the
tasks.  If you're using a newer version with the trashcan, emptying the
trashcan could help a little too.

--
Greenbone Networks GmbH
Neuer Graben 17, 49074 Osnabrueck, Germany | AG Osnabrueck, HR B 202460
Executive Directors: Lukas Grunwald, Dr. Jan-Oliver Wagner



More information about the Openvas-discuss mailing list