I have just briefly scanned the docs for Subversion. Is it true my only choices for database are either Berkely DB or flat file? If that is true, is Berkely DB as bad as they make it sound? The docs talk about how if your server crashes in the middle of a read then you have to do some sort of difficult restore. How likely is this to occur? Which storage do you all use?

In first I was using that BerklyDB for storing, but I had several crashes and needed to build a whole new structure for my project (with all the revisions losts) multiple times! Now with FSFS I did not experience any of this, so I will sincerelly recommend you to go for FSFS!

What's FSFS?, sounds interesting I use berkeley DB and never had a problem and do several checkins a day on several projects.

FSFS is their Flat File alternative to using the Berkely DB. I have five developers that would be connecting to the database, so I'm wondering if Berkely DB can "handle" that or not.

BlueCell, how many developers are connecting to your db?

Frans, were you the only one connecting to yours?

NickD wrote:

FSFS is their Flat File alternative to using the Berkely DB. I have five developers that would be connecting to the database, so I'm wondering if Berkely DB can "handle" that or not.

ah thanks

Frans, were you the only one connecting to yours?

Yes. Sometimes 2 persons, but most of the time one.

NickD wrote:

BlueCell, how many developers are connecting to your db?

Only one!

Also, please note that the subversion faq says:

The BerkeleyDB database in your repository is sensitive to interruptions. If a process accessing the database exits without "cleanly" closing the environment, then the database is left in an inconsistent state. Common causes of this include:

the process exiting when it hits a permission problem the process crashing/segfaulting the process being forcibly killed running out of disk space

However, sometimes, when you 'misuse' your subversion client and eg checkout wrongly, the occurs too! IMO this should not be the case (even if your doing something wrong). I found FSFS solved this.

Is it easy to change to fsfs ? (I didn't re-read the manual for this ) And does it give a lot of files/extra data or doesn't it make that much difference?

Have a look at the following comparison table: http://svnbook.red-bean.com/en/1.1/ch05.html#svn-ch-5-sect-1.3

I implemented FSFS using SVN Service: http://dark.clansoft.dk/~mbn/svnservice/

The installation was not hard at all, thought you need to make some changes in the config file. The actual pages (hosted on tigris.org) on how to do this are offline (host is offline), so I am afraid I can't point you to the right page describing this process.

Otis wrote:

Is it easy to change to fsfs ? (I didn't re-read the manual for this ) And does it give a lot of files/extra data or doesn't it make that much difference?

I switched from BerkeleyDB to FSFS and it was just a matter of backing up the BerkeleyDB and restoring it to a new repo that used FSFS. The backup and restore were done using the "standard" subversion backup and restore tools. (I can't recall off the top of my head what they are.)

The SVN book also has details about this: http://svnbook.red-bean.com/en/1.1/ch05s03.html#svn-ch-5-sect-3.5