[pgpool-general: 9412] Re: FATAL: failed to read kind from backend

Ron Johnson ronljohnsonjr at gmail.com
Thu Mar 27 23:00:38 JST 2025


On Thu, Mar 27, 2025 at 9:28 AM Tatsuo Ishii <ishii at postgresql.org> wrote:

> > Last week, I disabled load balancing, but the replica still has some
> weight
> > Might that be causing the problem?
> > load_balance_mode = off
> > backend_weight0 = 2
> > backend_weight1 = 1
>
> No, I don't think so. If load_balance_mode = off, then backend_weight
> is ignored.


I'm still getting the WARNING and FATAL messages even after stopping pgpool
on all three nodes, but only starting pgpool on nodes C and A.

pcp_watchdog_info thinks that node B is DEAD.

[postgres at FISPCDSPGS302c ~]$ pcp_watchdog_info -wv
Watchdog Cluster Information
Total Nodes              : 3
Remote Nodes             : 2
Member Remote Nodes      : 2
Alive Remote Nodes       : 1
Nodes required for quorum: 2
Quorum state             : QUORUM EXIST
Local node escalation    : YES
Leader Node Name         : FISPCDSPGS302c:9999 Linux FISPCDSPGS302c
Leader Host Name         : FISPCDSPGS302c

Watchdog Node Information
Node Name         : FISPCDSPGS302c:9999 Linux FISPCDSPGS302c
Host Name         : FISPCDSPGS302c
Delegate IP       : 10.109.160.20
Pgpool port       : 9999
Watchdog port     : 9000
Node priority     : 1
Status            : 4
Status Name       : LEADER
Membership Status : MEMBER

Node Name         : FISPCDSPGS302a:9999 Linux FISPCDSPGS302a
Host Name         : FISPCDSPGS302a
Delegate IP       : 10.109.160.20
Pgpool port       : 9999
Watchdog port     : 9000
Node priority     : 1
Status            : 7
Status Name       : STANDBY
Membership Status : MEMBER

Node Name         : Not_Set
Host Name         : FISPCDSPGS302b
Delegate IP       : Not_Set
Pgpool port       : 9999
Watchdog port     : 9000
Node priority     : 0
Status            : 0
Status Name       : DEAD
Membership Status : MEMBER

[postgres at FISPCDSPGS302c ~]$ pcp_health_check_stats -vw -n0
Node Id                       : 0
Host Name                     : FISPCDSPGS302a
Port                          : 5432
Status                        : up
Role                          : primary
Last Status Change            : 2025-03-27 08:49:59
Total Count                   : 766
Success Count                 : 766
Fail Count                    : 0
Skip Count                    : 0
Retry Count                   : 0
Average Retry Count           : 0.000000
Max Retry Count               : 0
Max Health Check Duration     : 17
Minimum Health Check Duration : 10
Average Health Check Duration : 11.481723
Last Health Check             : 2025-03-27 09:53:14
Last Successful Health Check  : 2025-03-27 09:53:14
Last Skip Health Check        :
Last Failed Health Check      :

[postgres at FISPCDSPGS302c ~]$ pcp_health_check_stats -vw -n1
Node Id                       : 1
Host Name                     : FISPCDSPGS302b
Port                          : 5432
Status                        : up
Role                          : standby
Last Status Change            : 2025-03-27 08:49:59
Total Count                   : 768
Success Count                 : 768
Fail Count                    : 0
Skip Count                    : 0
Retry Count                   : 0
Average Retry Count           : 0.000000
Max Retry Count               : 0
Max Health Check Duration     : 16
Minimum Health Check Duration : 10
Average Health Check Duration : 11.330729
Last Health Check             : 2025-03-27 09:53:24
Last Successful Health Check  : 2025-03-27 09:53:24
Last Skip Health Check        :
Last Failed Health Check      :


-- 
Death to <Redacted>, and butter sauce.
Don't boil me, I'm still alive.
<Redacted> lobster!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.pgpool.net/pipermail/pgpool-general/attachments/20250327/e8b9d310/attachment.htm>


More information about the pgpool-general mailing list