Mani's Blog

May 19, 2015

LOG: restarted WAL streaming at %S on timeline 1

Filed under: Uncategorized — mani @ 4:30 am

Problem:

I had setup like this using streaming replication.  Once I promote Slave1 as Master, the replication on Slave2 is broken with following error messages.

Master –> Slave1 –> Slave2

2015-05-19_01:00:19.55306 2015-05-19 01:00:19.553 UTC: @/555a7720.599a[2986]: LOG: restarted WAL streaming at 449/1E000000 on timeline 1
2015-05-19_01:00:19.55317 2015-05-19 01:00:19.553 UTC: @/555a7720.599a[2987]: LOG: replication terminated by primary server
2015-05-19_01:00:19.55319 2015-05-19 01:00:19.553 UTC: @/555a7720.599a[2988]: DETAIL: End of WAL reached on timeline 1 at 449/1E000090.

Solution:

I had to copy files from pg_xlogs directory from Master to Slave box under pg_xlogs directory.   Added recovery_target_timeline = ‘latest’  in recovery.conf, then restarted the slave DB.

Blog at WordPress.com.