Re: Postgresql HA questions

[Israel Brewster <israel@xxxxxxxxxxxxxx>]

Thanks. I'd taken some stabs at setting up Corosync/Pacemaker, but could never get the PostgreSQL portion to work properly. One difficulty is that I am using named replication slots, which don't appear to be supported. The other issue is that the system tended to try to start my secondary as primary, which of course doesn't work to bring up the old primary as secondary (at least, that's my understanding. Let me know if that's wrong).

I'll take another stab at it given the steps outlined in the presentation you posted, and given your success - I must just be doing something wrong here. 

-----------------------------------------------

Israel Brewster

Systems Analyst II

Ravn Alaska

5245 Airport Industrial Rd

Fairbanks, AK 99709

(907) 450-7293

-----------------------------------------------

BEGIN:VCARD
VERSION:3.0
N:Brewster;Israel;;;
FN:Israel Brewster
ORG:Frontier Flying Service;MIS
TITLE:PC Support Tech II
EMAIL;type=INTERNET;type=WORK;type=pref:israel@xxxxxxxxxxxxxxxxxx
TEL;type=WORK;type=pref:907-450-7293
item1.ADR;type=WORK;type=pref:;;5245 Airport Industrial Wy;Fairbanks;AK;99701;
item1.X-ABADR:us
CATEGORIES:General
X-ABUID:36305438-95EA-4410-91AB-45D16CABCDDC\:ABPerson
END:VCARD

On Sep 25, 2015, at 1:46 AM, Steve Pritchard <steve.pritchard@xxxxxxx> wrote:

Israel,

I can't answer all your questions, but we've just set up a HA pair with Hot Standby using Corosync/Pacemaker. However we haven't deployed this 'live' yet. 

We originally found a presentation from The PostgreSQL Conference PostgreSQL High Availability with Corosync/Pacemaker, and then bought the book PostgreSQL 9.0 High Performance (a sample chapter is available as a PDF).

All working out well in testing at the moment.

Steve Pritchard

British Trust for Ornithology

On 23 September 2015 at 17:36, Israel Brewster <israel@xxxxxxxxxxxxxx> wrote:

<snip>

With my application servers, I have a system set up using corosync and pacemaker that allows for seamless fail-over between the two machines, with the IP address and all services moving smoothly between the two at will. Ideally, I would have a similar setup with my database servers, so the applications never even know that there was a switch. Is this possible with Postgresql at all? Does it make a difference that at least one app has an "always on" connection to the DB Server?

<snip>