From: | Fujii Masao <masao(dot)fujii(at)gmail(dot)com> |
---|---|
To: | Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com> |
Cc: | PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: logical rep worker for table sync can start and stop very frequently unexpectedly |
Date: | 2017-04-12 13:57:10 |
Message-ID: | CAHGQGwE6oBYFBOSeB40rcL7g=LSXaEHYQHZ-Td_b66eJjF3DYw@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On Wed, Apr 12, 2017 at 2:40 AM, Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com> wrote:
> On Wed, Apr 12, 2017 at 2:05 AM, Fujii Masao <masao(dot)fujii(at)gmail(dot)com> wrote:
>> Hi,
>>
>> I observed $subject when I ran the following steps.
>>
>> 1. start the publisher server
>> 2. start the subscriber server
>> 3. create table with primary key and publication for it
>> 4. insert several records into the table
>> 5. create table with primary key and subscription
>> 6. wait for table sync to finish
>> 7. drop the subscription and recreate the subscription
>>
>> Then the launcher starts the worker, and the worker starts
>> another worker for table sync. But that worker for table sync
>> exits immediately because of primary key violation.
>> Then the worker tries to start new worker for table sync immediately
>> and it also exits immediately because of primary key violation.
>> This sequence repeats very fast...
>>
>> Attached is the script that I used to reproduce the issue.
>>
>
> I encountered the same situation[1] and am proposing to put interval
> of launching table sync worker after failed to attempt.
>
> [1] https://www.postgresql.org/message-id/CAD21AoDCnyRJDUY%3DESVVe68AukvOP2dFomTeBFpAd1TiFbjsGg%40mail.gmail.com
Okay, let's discuss this in that thread. Thanks!
Regards,
--
Fujii Masao
From | Date | Subject | |
---|---|---|---|
Next Message | Alexander Kuzmenkov | 2017-04-12 14:01:39 | Re: index-only count(*) for indexes supporting bitmap scans |
Previous Message | Fujii Masao | 2017-04-12 13:55:08 | logical replication and PANIC during shutdown checkpoint in publisher |