I have three Database servers, running PostgreSQL 9.2:
master01 (Own data center USA)
slave01 (Own data center USA)
slave02 (AWS Australia)
Both slaves use streaming replication + wal_files (delivered by master01 via ssh).
The postgres process on the slave02 server has been recently stopped, due a maintenance on the slave02's postgres disk volume(
Postgres is running on:
Wal files are shipped to this volume:
Once the volume was restored, I started postgres and got this log:
cp: cannot stat '/var/lib/walfiles/0000000500008AA800000069': No such file or directory cp: cannot stat '/var/lib/walfiles/0000000500008AA80000006A': No such file or directory consistent recovery state reached at 8AA8/8AD22F70 record with zero length at 8AA8/8AD22F70 database system is ready to accept read only connections streaming replication successfully connected to primary
I have a very large table, that is constantly updated. I ran this query on slave01 and slave02 server:
select max(id) from ids_table;
max ---------- 45148099 (1 row)
max ---------- 45148095 (1 row)
Can you see the results are different? And that shouldn't be different!
How can I check the data between servers? How can I see if there is something wrong with slave02 due to those
No such file or directory errors?
As far as I could see, checksum is only available on Postgres 9.3+ ? So can't use that!
UPDATE 2: Anwsering Sahap Asci's questions
On the master:
synchronous_commit = on #synchronous_standby_names = ''