Postgresql and shared memory
I tried running PostgreSQL as one of my first experiments, and hit upon some nuances of shared memory, which may be helpful to relate here.
One of the easiest ways to increase Postgres's performance is to turn up the shared_buffers parameter in the postgresql.conf file; this is basically the amount of shared memory which the postmaster will use for buffering everything: table data, indexes, etc. The default value is really small, and if you have RAM to spare you may want to crank it up to 128MB (that's 16384 shared buffers, for pgsql 8.1 and earlier) or even more. But if it then complains that it couldn't allocate the shared memory, this page explains why.
PostgreSQL uses shared memory. As such, you have to check two things about the VE.
1. UBC's shmpages setting for this VE. This dictates how many pages (8k apiece) are available to the VE, e.g. shmpages=16384 gives a limit of 128 MB of shared memory.
2. "/sbin/sysctl kernel.shmmax" This is the VE's self-imposed limit on how much shared memory may be allocated in a single request.
Get that? Not only does the HN impose a limit on the VE's total shared memory usage, but the VE itself has a setting for the maximum amount in a single chunk.
Some other notes:
- Note that shared memory is part of the VE's overall memory usage. It is not a second memory pool.
- Note that processes other than PostgreSQL may be using shared memory, and also that shmpages includes stuff other than IPC shared memory (tmpfs, shmem, etc) So don't set shmpages to exactly the amount you want to give Postgres.
- The sysctl kernel.shmmax value set in the HN/VE0 applies only to the HN, not to VEs.
