[pgpool-general: 6984] Re: Failover fails to start delegate IP

[Tatsuo Ishii]

No problem. I have added a FAQ entry:

https://pgpool.net/mediawiki/index.php/FAQ#Watchdog_standby_does_not_take_over_master_when_it_goes_down

> Yeah I think you had told me that before but forgot about that - sorry
> 
> thx
> 
> On 4/14/20, 7:16 PM, "Tatsuo Ishii" <ishii at sraoss.co.jp> wrote:
> 
>     > I have 2 pgpool nodes - primary and standby using watchdog. When I stop pgpool on the primary to failover to the standby the delegate IP never starts. In the pgpool.log file on the standby I see
>     > “I am the cluster leader node but we do not have enough nodes in cluster”
>     > Now if a start pgpool on the primary, the delegate IP starts on the standby.  At the point the standby has the delegate IP, both nodes are up running pgpool. So I stop pgpool on the standby. Now the primary takes over and has the delegate IP.
>     > 
>     > So failing over to the standby, fails to get the delegate IP until the primary has pgpool running again. But the standby will failover to the primary and get the delegate IP without pgpool running on the standby.
>     > 
>     > Host OS = Ubuntu 18.04
>     > PgPool version = 4.1.1 and tried on 4.1.0 same issue.
>     > 
>     > If I do the same test using PgPool version 4.0.5 the failover works both ways getting the delegate IP
>     
>     You need to trun on enable_consensus_with_half_votes. See 4.1 release
>     note for more details.
>     https://www.pgpool.net/docs/latest/en/html/release-4-1-0.html
>     
>     Best regards,
>     --
>     Tatsuo Ishii
>     SRA OSS, Inc. Japan
>     English: http://www.sraoss.co.jp/index_en.php
>     Japanese:http://www.sraoss.co.jp
>     
>