aboutsummaryrefslogtreecommitdiff
path: root/src/tutorial/syscat.source
diff options
context:
space:
mode:
authorKevin Grittner <kgrittn@postgresql.org>2016-04-12 11:48:02 -0500
committerKevin Grittner <kgrittn@postgresql.org>2016-04-12 11:48:02 -0500
commit2201d801b03c2d1b0bce4d6580b718dc34d38b3e (patch)
treea656c7e2e8349ebae3ed3bd8edf101de7a245a9f /src/tutorial/syscat.source
parent5713f03973e26ad6df6df5ac8b9efa0123d68062 (diff)
downloadpostgresql-2201d801b03c2d1b0bce4d6580b718dc34d38b3e.tar.gz
postgresql-2201d801b03c2d1b0bce4d6580b718dc34d38b3e.zip
Avoid extra locks in GetSnapshotData if old_snapshot_threshold < 0
On a big NUMA machine with 1000 connections in saturation load there was a performance regression due to spinlock contention, for acquiring values which were never used. Just fill with dummy values if we're not going to use them. This patch has not been benchmarked yet on a big NUMA machine, but it seems like a good idea on general principle, and it seemed to prevent an apparent 2.2% regression on a single-socket i7 box running 200 connections at saturation load.
Diffstat (limited to 'src/tutorial/syscat.source')
0 files changed, 0 insertions, 0 deletions