Zitat von "Luis H. Forchesatto" <luisforchesatto AT gmail DOT com>:

> Greetings.
>
> I'd like to discuss my situation here where I have a job who backups only
> 200GB in files, but it has 2 million more files to save on catalog (MySQL).
> When the file copy is complete the storage server, who also runs the
> director, spend many hours to save the files in the catalog. This also
> cause low mysqld performance at all.
>
> Is there any tip to optimize the catalog operation or MySQL performance to
> make the job less resource hungry?
>
> The job runs once a week with no concurrent jobs. Director and storage runs
> at the same computer but the client is another server, at the same network
> who can transport up to 1Gb of data between the servers.
>
> Any tips will be apretiated.

Maybe the wrong database? Our main filer has ~2.7 million files per  
full backup which pile up to ~1GB attribute data to go in the backup  
database (PostgreSQL). It finish to despool the attributes within 23  
minutes on a commodity hardware. If you have to stick with MySQL use  
Bacula attribute spooling feature and tune MySQL for fastest insert  
with something along these general tips:

- Raise the size of the redologs
- Enlarge the buffers
- Use a really fast dedicated i/o channel for the db

Some further reading is here:

http://stackoverflow.com/questions/7585301/how-tos-for-mysql-innodb-insert-performance-optimization

Regards

Andreas




------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users