A while ago I wrote a small script which runs benchmarks against given filesystems and collects performance data for each run. What I wanted to find out is, how expensive (IO-wise) various standard filesystem operations are.
The collected informations proved to be quite extensive and very hard to visualize.
I'll be giving a workshop on PostgreSQL replication at the Netways Open Source Datacenter Conference in Nuremberg, Germany on 29. & 30. April 2009.
Topics will include a short introduction in the theory of data replication in RDBMS context as well as giving an overview of the various replication solutions which are currently available for PostgreSQL. The hands-on part will focus on a few common scenarios with Log Shipping and Slony-I (TBD).
The workshop has only 20 slots available, so register early if you're interested.
