On 11/4/2013 12:15 PM, Phil Stracchino wrote:
> On 11/04/13 09:58, compdoc wrote:
>> By the way, it wasn't enough to enable InnoDB - I had to create the bacula
>> database after it was enabled for the tables to use this engine. (it was a
>> new install)
>>
>> I don't know if it's possible to convert the tables after enabling InnoDB,
>> but I would think it's better to start them out that way. You can google
>> ways to tell if your tables are using InnoDB.
>
> It is trivial to convert the tables to InnoDB after creation.  Here's a
> simple Unix bash shell one-liner to do it:
>
> for T in $(mysql -BNe 'show tables from bacula'); do echo Converting
> table bacula.${T} ; mysql -BNe "alter table bacula.${T} engine=InnoDB" ;
> done
>
> Any time you're working with InnoDB or converting tables to InnoDB,
> remember that within limits, the more memory you have allocated to the
> InnoDB buffer pool, the faster it will go.

I would add that it is critical (IMO) to place the DB storage on 
different physical drives than those holding the Bacula spool area. At 
the end of a job Bacula SD must read the spooled attributes and update 
the catalog. If spooled attributes and catalog are on the same spindles, 
then the disk thrashing negates the advantage of attribute spooling. I 
might also add that placing the catalog on SSD resulted in a hefty 
performance gain.

As for clustering using DRBD, the catalog and spool area should still be 
on different spindles.


------------------------------------------------------------------------------
Android is increasing in popularity, but the open development platform that
developers love is also attractive to malware creators. Download this white
paper to learn more about secure code signing practices that can help keep
Android apps secure.
http://pubads.g.doubleclick.net/gampad/clk?id=65839951&iu=/4140/ostg.clktrk
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users