On Jul 4, 2005, at 8:31 PM, Adrian Hoe wrote:
>
> But reliability is more then easy integration/setup. We have
> systems running on MySQL with more than 250 tables and over 7
> millions tuples and not crashing at all.
Are you aware that mysql uses BerkeleyDB for transacations? :-)
Actually, your mysql database probably has crashed many times, but
you don't notice. That's because the 'mysqld' process is a single
point of access for all programs; if the database gets inconsistent,
mysqld stops serving all requests, recovers the database, then things
move on smoothly.
The big problem between svn and BDB is that we didn't do it this
way: there's no single 'svnd' mediating all access to the tables and
keeping guard over database health. That's the model that BerkeleyDB
expected we'd follow, and we didn't. Instead, we have N diferent
programs all using the database (via library linkage), and when one
program wedges the database, all the others sit around and get stuck
waiting... or error out. Rather than a daemon automatically noticing
the problem and recovering, a human has to step in and recover things.
So our big mistake here was not realizing how 'brittle' (easy to
wedge) BDB would be, in the absence of a central daemon.
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Tue Jul 5 03:55:23 2005