[pgpool-general: 6785] Re: VIP and quorum

[Tatsuo Ishii]

Hi Usama,

Can you please take a look at this?

Best regards,
--
Tatsuo Ishii
SRA OSS, Inc. Japan
English: http://www.sraoss.co.jp/index_en.php
Japanese:http://www.sraoss.co.jp

> Hi all,
> 
> I am testing pgpool 4.1.0 and ended up in something I found inconsistent.
> This is the description of the delegate_IP parameter in 4.1:
> 
>> Specifies the virtual IP address (VIP) of Pgpool-II that is connected
> from client servers (application servers etc.). When a Pgpool-II is
> switched from standby to active, the Pgpool-II takes over this VIP. If
> failover_require_consensus
> <https://www.pgpool.net/docs/latest/en/html/runtime-watchdog-config.html#GUC-FAILOVER-REQUIRE-CONSENSUS>
> is
> on (the default), VIP will not be brought up in case the quorum does not
> exist. Default is ''(empty): which means virtual IP will never be brought
> up.
> 
> The part that explains about failover_require_consensus says that if this
> parameter is enabled then VIP will not be claimed if quorum does NOT exist.
> I assume that if I DISABLE this parameter then VIP will be claimed even
> there is no quorum.
> 
> However, when testing with a 3 nodes cluster and failover_require_consensus
> disabled, when I shutdown two nodes the third one doesn't claim the VIP.
> This is what the log states:
> 
>> LOG:  I am the cluster leader node but we do not have enough nodes in
> cluster
>> DETAIL:  waiting for the quorum to start escalation process
> 
> In fact, I checked the source code and found no relationship between
> failover_require_consensus
> and escalation.
> 
> I'm not complaining about the behavior, just that it seems inconsistent
> with what the doc states. Or perhaps I misunderstood it.
> 
> Regards,
> 
> Rafael Castro.