Hi
I am having some problem understanding the details of a couple of
configuration directives. Here is what I think I understand it so far:
effective_cache_size
- specifies in how many B/KB/MB will be used to cache data
between pg and the OS?
- what is actually cached and how it is actually used by the
planner and how does it affect the planner and the queries?
-in other words, when and why do I need to set this parameter?
checkpoint_segments
- specifies the number of segments?
size: number*16MB?
- means when number of WAL segments has filled up and matches
checkpoint segment, dirty data is written to disk?
- so it specifies how much data is stored in memory/wal before
written to disk and therefore affects recoverability?
regards
thomas