This post originated from an RSS feed registered with Python Buzz
by Phillip Pearson.
Original Post: Subversion: better the second time around
Feed Title: Second p0st
Feed URL: http://www.myelin.co.nz/post/rss.xml
Feed Description: Tech notes and web hackery from the guy that brought you bzero, Python Community Server, the Blogging Ecosystem and the Internet Topic Exchange
Last time I tried using Subversion, I installed a 1.0 version on my Debian/woody server, running svnserve behind SSH -- pretty much the same way as I run CVS. This is nice if you already have the infrastructure to run CVS like this, as it lets you use SSH keys for authentication, and SSH keys are really convenient.
However, within a day of starting using it, I found that some checkins were causing the server to grind to a near-halt; svnserve was using several hundred megabytes of memory, and the machine was spending all its time swapping. So I went back to CVS.
Giving Subversion another go now, I've installed the 1.1 version and Apache 2 on a new Debian/sarge server, and now have SSL running (that took ages), so everything runs over HTTPS via Apache rather than SSH and svnserve. It seems happier. This server has twice as much RAM - 512 MB - which may be making a difference, but it seems that either 1.1 or the Apache/DAV route makes things a bit more stable. It's also an fsfs database rather than a BSDDB one.
So: let's see how this goes. Looks good so far; svn rename has already come in handy.
Previous Topic
