[dev.icinga.com #10969] IDO tries to failover even when only one of two masters has IDO configured

[icinga-migration]

This issue has been migrated from Redmine: https://dev.icinga.com/issues/10969

Created by twidhalm on 2016-01-14 09:20:53 +00:00

Assignee: (none)
Status: Rejected (closed on 2016-01-14 09:31:11 +00:00)
Target Version: (none)
Last Update: 2016-01-14 09:31:11 +00:00 (in Redmine)

Icinga Version: 2.4.1
Backport?: Not yet backported
Include in Changelog: 1

---

Hi,

I have a "minimal HA" setup, where only one Icinga 2 Master was planned. For redundancy a second master was introduced, but only with a minimal setup: Only Icinga 2, no IDO, no Grapher, etc. . In fact this second master was mostly created to send notifications while the main master ist down.

When I start the second master, die IDO connection on the main master goes down and "Pausing IDO connection" is logged. When I set enable_ha to false it works like expected.

I think it would be more logical if failover / ha is only active when other masters have an IDO connection configured as well not just when another master is there.

Cheers,
Thomas

[icinga-migration]

Updated by mfriedrich on 2016-01-14 09:31:11 +00:00

• Category set to Cluster
• Status changed from New to Rejected

That's not supported. All features providing HA capabilities must remain enabled on all endpoints in a zone.

http://docs.icinga.org/icinga2/latest/doc/module/icinga2/chapter/distributed-monitoring-high-availability#high-availability-features

All nodes in the same zone require the same features enabled for High Availability (HA) amongst them.

Kind regards,
Michael