This is a public chat log generated from the #semsol IRC channel.
14:09:17
well, in 24 hours i've loaded 14M triples. In the meantime, I extracted the triples that I need to solve a short-term problem with a grep pipeline in about 2 min.
14:10:12
at this point i'm blaming mysql and my disk subsytem more than I'm blaming arc. I'm thinking seriously about attaching a high performance drive to the machine for this kind of work
14:10:35
also wondering if i can drop some of the indexes on arc's tables for the load, them build them later. arc creates a lot of indexes!
14:23:06
this morning it looks like mysqld cpu usage is the bottleneck. i'm seeing big i/o bursts, but atop sees mysql cpu usage between 50-105% now
14:23:17
sometimes i/o drops below 30%
18:52:15
i ran into some trouble here... all of the id's are medium ints, so i hit a wall at 16M tuples
18:52:36
i need to see if the database is usable at 16M tuples, and if it is, change the identifiers... dan
18:52:42
(i mean damn...)