[pgpool-general: 4802] Re: 3 or more postgres backends and failover.
ishii at postgresql.org
Thu Jul 21 17:07:41 JST 2016
>> Because pgpool-II does not automatically promote standby, when fail
>> over gets called, new primary is not defined yet. It is the fail over
>> script's responsibility to promote a standby.
> Ah, i see. As i understood, pgpool makes a suggestion which standby
> should be promoted, but which one would actually be promoted is
> determined by failover script. Thanks.
>>> Ok, youngest node id makes sense. Thanks. But why does it needs to
>>> start with 0? As i understand, its an arbitrary number set in config
>>> file, but when i set backends starting with "backend1_*", i've got
>>> empty node with node_id 0 in "show pool_nodes".
>> That's how pgpool-II is implemented. Do you have any specific reason
>> to want to start with backend1_*, rather than backend_0*?
> Not as such, no. Its just when i use pgpool in conjunction with repmgr
> (which also uses arbitrary node ids), i want to use the same ids and not
> always start with 0. Just a matter of good looks.
Intersting. repmgr wants to start from 1, rather than 0, I guess.
SRA OSS, Inc. Japan
More information about the pgpool-general