On 17 May 2015 at 04:14, Mathew Moon <mathew.moon@vipaar.com> wrote:
> Hi,
>
> Wanted to "bump" this (Is that ok in this list?). Like I mentioned, I see
> now that the error I am seeing in the log possibly indicates that the node I
> am "reinitializing" is creating a slot that points to its own old sysid.
> Checking manually I see that this slot does in fact exist. I still have no
> idea why it is creating this incorrect slot and trying to connect to it. If
> anyone has any insight into this it would be greatly appreciated.
I fixed a bug just after the removal of 0.9.2 that is the most likely
explanation for this. BDR wasn't correctly ignoring
bdr.bdr_connections entries for 'k'illed nodes when joining a new
node.
> Also, does anyone have experience using repmgr for streaming replication
> (Log shipping standby)? If so what do you think about it?
2ndQuadrant maintains repmgr. I think it's a decent tool for the job.
> Is it a good tool
> or is there something better for the same use case that repmgr fulfills.
It's moderately easy to just do it by hand if you don't want to use
repmgr. I haven't looked for alternative tools for this use case.
--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
