BUG #14837: pglogical subscription strange behavior

From: gkostov79(at)gmail(dot)com
To: pgsql-bugs(at)postgresql(dot)org
Cc: gkostov79(at)gmail(dot)com
Subject: BUG #14837: pglogical subscription strange behavior
Date: 2017-09-29 09:52:06
Message-ID: 20170929095206.28558.87966@wrigleys.postgresql.org
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

The following bug has been logged on the website:

Bug reference: 14837
Logged by: Georgi Kostov
Email address: gkostov79(at)gmail(dot)com
PostgreSQL version: 9.6.5
Operating system: Debian
Description:

Hi guys,
I have bit strange problem with pglogical subscription: at first look on
randomly after "select pglogical.alter_subscription_disable('$dbs',false) "
and then select pglogical.alter_subscription_enable('$dbs',false)" not all
subscriptions are replicating this happen rarely but if happen subscription
stay down and not replicating until cluster restart. after that everthing is
fine and all are in state replicating.
In the logs I have only this :
2017-09-29 09:30:26.155 UTC,,,27322,,59ce12b2.6aba,1,,2017-09-29 09:30:26
UTC,5/0,0,LOG,00000,"manager worker [27322] at slot 3 generation 70
detaching cleanly",,,,,,,,,"pglogical manager 12407"
2017-09-29 09:30:26.157 UTC,,,27323,,59ce12b2.6abb,1,,2017-09-29 09:30:26
UTC,5/0,0,LOG,00000,"manager worker [27323] at slot 3 generation 71
detaching cleanly",,,,,,,,,"pglogical manager 1"
Can someone advice me what to do to avoid cluster restart please?

Browse pgsql-bugs by date

  From Date Subject
Next Message zam6ak 2017-09-29 13:43:21 BUG #14838: clock_timestamp() returns same values for each row
Previous Message aistrong.hs 2017-09-29 09:28:41 BUG #14836: Is there an option to change and run isolation level only for a single statement as in DB2