[Errno 8] nodename nor servname provided, or not known

[Matter]

Since approximately four weeks we see a lot of clients with the above error in MunkiReport. Normally there are always a couple of clients with this error, but now we see up to 10% of the clients showing it. It seems that in the morning the number of errors are high and during the day it goes down. This repeats itself every day. At first it was only one specific environment, but now more environments are showing the same error.

To me it looks like some sort of network issue, maybe it's to crowded or something. I have done a ping session to the munki server during three hours (9000 pings) and got 2 timeouts in total. Seems quiet normal. Also I have a monitoring system in place which monitors the munki_repo and munkireport on availability. I'm not getting any alerts from this system, also is the service not flapping.

I'm looking for more ways to troubleshoot to find out the root cause.

We're on Munki 2.8.0.2810

Errors are seen on multiple OS versions (10.8 - 10.11)

[Gregory Neagle]

This is a DNS resolution error.

A guess: Munki is running soon after wake/boot, but before DNS resolution is working. The serve check fails, and Munki proceeds to running the postflight. Your postflight is either configured to submit MR data to a URL containing an IP address, or by the time the postflight executes, DNS resolution is working.

-Greg

--
Find related discussion groups here:
https://github.com/munki/munki/wiki/Discussion-Group
---
You received this message because you are subscribed to the Google Groups "munki-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to munki-dev+...@googlegroups.com.
To post to this group, send email to munk...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Matter]

Just found out that MunkiReport 2.9 runs the reporting code on preflight (before the actual munki run), so it actually reports on the previous munki run. And indeed, a closer look shows that the errors are mostly from the day before. So the errors occur at the end of the day.

However, if it's indeed a DNS issue I'll temporary change the SoftwareRepoURL to the IP address. Lets see what happens.

Op dinsdag 11 oktober 2016 17:44:36 UTC+2 schreef gregn...@mac.com:

[Eric Holtam]

This commit on the wip branch of MunkiReport moves the munki reporting to the postflight.  After testing with IP maybe try the wip branch, too.

-Eric

[Gregory Neagle]

On Oct 12, 2016, at 2:03 AM, Matter <matti...@gmail.com> wrote:

Just found out that MunkiReport 2.9 runs the reporting code on preflight (before the actual munki run), so it actually reports on the previous munki run.

So a machine that could not reach your Munki server (a laptop taken off the network, for example) would then report that issue the next time the MunkiReport preflight script ran.

I think we can end discussion of this topic here (as there's nothing related to the development of Munki), and if you are concerned/bothered/want to fix this, you should begin discussions with the MunkiReport developers.

-Greg

[Matter]

That's what I did in the first place. That's when I learned that the reporting code ran on preflight. However, I'm sure that a number of workstations have not been off the network when this error occurred. So that's why I posted it anyway to find more ways to troubleshoot. But I guess the number of errors will decrease a lot when MunkiReport is updated. 

Op woensdag 12 oktober 2016 15:59:24 UTC+2 schreef gregn...@mac.com: