Logs files, Check points and recovery... oh my...

[pk...@connected.com]

Hi,  I recently went through a recovery and it took about 8 hours (and
perhaps several years off of my life).  In my repository I have about 2.5 GB
of unused log files and it appears that no one has used the BerkeleyDB
utility db_checkpoint to make a fall back point.  Do any of you make
checkpoints?  I see by looking at the trail code that svn should be doing
this on its own, but I wonder if there isn't a difference between this and
the utilitiy...

According to the BerkeleyDB pages:
http://pybsddb.sourceforge.net/ref/transapp/checkpoint.html

> Performing checkpoints is necessary for two reasons. First, you can remove

> the Berkeley DB log files from your system only after a checkpoint.
Second, 
> the frequency of your checkpoints is inversely proportional to the amount
of time
> it takes to run database recovery after a system or application failure. 

Does anyone know if a repository administrator should be using the
db_checkpoint utility
(http://pybsddb.sourceforge.net/utility/db_checkpoint.html)?

Also, I want to test wedging the db and recovery to prove to myself which
checkpoint method is the best for my repository.  What's the best way to
wedge a db?  Make a large commit and kill the serving process in the midst
of it?

Thanks for the help....

Re: Logs files, Check points and recovery... oh my...

[Bastian Blank <ba...@waldi.eu.org>]

On Sat, Sep 04, 2004 at 11:10:26PM -0400, pkahn@connected.com wrote:
> Also, I want to test wedging the db and recovery to prove to myself which
> checkpoint method is the best for my repository.  What's the best way to
> wedge a db?  Make a large commit and kill the serving process in the midst
> of it?

100 simultane checkouts.

Bastian

-- 
It would be illogical to kill without reason.
		-- Spock, "Journey to Babel", stardate 3842.4

Re: Logs files, Check points and recovery... oh my...

[Max Bowsher <ma...@ukf.net>]

pkahn@connected.com wrote:
> Hi,  I recently went through a recovery and it took about 8 hours (and
> perhaps several years off of my life).  In my repository I have about 2.5 
> GB
> of unused log files and it appears that no one has used the BerkeleyDB
> utility db_checkpoint to make a fall back point.  Do any of you make
> checkpoints?  I see by looking at the trail code that svn should be doing
> this on its own, but I wonder if there isn't a difference between this and
> the utilitiy...

No difference whatsoever.

You BDB is broken if it is not doing checkpoints when svn asks it to.


Max.


---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org