AFAIK it helps at least bulk loading my data every other time.
So I'm confused and backup again: Given a single-disk virtual Linux
system and a 'read-only' dataset, which is exposed to the internet and
completely replaced from time to time, and expecting SELECT queries
including joins, sorts, equality and range (sub-)queries...
=> What are the suggested postgresql.conf and session parameters for
such a "read-only database" to "Whac-A-Mole" (i.e. to consider :->)?
Stefan
2011/4/23 Robert Haas <robertmhaas@gmail.com>:
> On Apr 18, 2011, at 6:08 PM, Stefan Keller <sfkeller@gmail.com> wrote:
>> I browsed the faq and looked at PostgreSQL performance books but I
>> could not find the obvious:
>> How to configure a read-only database server?
>>
>> I have a single-disk virtual Linux system and a read-only dataset
>> which is exposed to internet and completely replaced from time to
>> time.
>>
>> This is what I found so far:
>>
>> * Disabling autovacuum daemon.
>> * Setting postgresql.conf parameters:
>> fsync=off
>> synchronous_commit=off
>> full_page_writes=off
>
> All of those speed up writes. I don't know that they will make any difference at all on a read-only workload.
>
>> * What about wal_level and archive_mode?
>
> Same with these.
>
>>
>
> ...Robert