Hello all, I have been spending the last few weeks trying out different methods for loading a batch of around 100,000 records. I have had varying degrees of success with getting the load to "work," including copy level information, but the last few loads seem to have somehow developed a search speed problem. I say "developed" because I am pretty sure that it wasn't horribly slow the first few times I did the load, but has gotten progressively worse to the point that any keyword search which returns several thousand results no longer completes in the OPAC, and takes from 45 seconds to several minutes to process in postgres. I am not certain that it has gotten worse, but I am sure that it is currently quite bad.
In between each load I have been running build-db.sh, and even dropped and recreated my entire evergreen database before a recent load. Is there something else I need to do in order to get an entirely clean slate? This project has been my first experience with postgres, so I may be missing something rather obvious. My server specs are somewhat modest but by no means slow; I am currently running a RedHat ES 5 virtual machine on an Opteron 2.4Ghz with 1GB of dedicated RAM. Any suggestions? Naturally one might suspect an index problem. Is there any way to verify that all relevant indexes are working, and/or to simply rebuild them all? Thanks, DW
