Multiple deploys at the same time will cause Comments/scheduled downtime loss

Hello,

We have found that running 2 deploys simultaneously on director will cause comments and scheduled downtimes to disappear.

version: 2.13.1-1
os: CentOS Linux release 7.9.2009
enabled feature list: api checker command compatlog graphite ido-mysql livestatus mainlog notification statusdata
Icinga Web 2 Version: 2.9.0, modules: setup 2.9.0, aws 1.1.0, ca 1.0.2, cube 1.1.1, director 1.8.0, elasticsearch 1.0.0, fileshipper 1.1.0, graphite 1.1.0, idoreports 0.9.1, incubator 0.6.0, ipl v0.5.0, jira 1.0.1, monitoring 2.9.0, nagvis 1.1.1, pdfexport 0.9.1, reactbundle 0.9.0, reporting 0.10.0

What use case do you have for multiple deploys? Just asking since I haven’t run across it.

Recently having a borked API package that only seemed to affect scheduled downtimes (not comments, acknowledgements, etc…) I would think that both deploys running simultaneously conflict with each other when trying to create an API package.

/var/lib/icinga2/api/packages/_api ← files created by the Director (and presumably other API deployments) go here, so I wouldn’t think that 2 Director deployments writing files here at the same time and performing all of it’s checks and balances would be safe.

I could be wrong, this is just a guess based off a recent experience that had a similar symptom.

Maybe you give us more further information? Are the two deployments on the same server or two diffrent master servers?
If it is on one server: is one deployment on a pending-like state?

Hello all!

Let’s continue discussion on GitHub:

Best,
AK

1 Like