Postgres Pro
Downloads
Home   >   mailing lists

Re: [HACKERS] logical replication - still unstable after all these months - Mailing list pgsql-hackers

From Jeff Janes
Subject Re: [HACKERS] logical replication - still unstable after all these months
Date
Msg-id CAMkU=1whG4nR726-vEU25m_uhWf60LpPh0riMQXVXhcjFL2jyA@mail.gmail.com
Whole thread Raw
In response to Re: [HACKERS] logical replication - still unstable after all thesemonths  (tushar <tushar.ahuja@enterprisedb.com>)
List pgsql-hackers
Tree view
On Fri, May 26, 2017 at 5:17 AM, tushar <tushar.ahuja@enterprisedb.com> wrote:

run second time =
./pgbench -T 20   -c 90 -j 90  -f test.sql  postgres

check the row count on master/standby
Master=
postgres=# select count(*) from pgbench_history ;
 count
--------
 536836
(1 row)

Standby =

postgres=#  select count(*) from pgbench_history ;
  count
---------
 1090959
(1 row)

Hi Tushar,

pgbench starts out by truncating pgbench_history.  That truncation does not get replicated to the subscriber.  The later inserts do.  So your subscriber ends up with about twice as many rows.

Cheers,

Jeff

pgsql-hackers by date:

Previous
From: Amit Kapila
Date:
Subject: Re: [HACKERS] Broken hint bits (freeze)
Next
From: "Regina Obe"
Date:
Subject: Re: [HACKERS] PostgreSQL 10 changes in exclusion constraints - did something change? CASE WHEN behavior oddity