"Site overview" in "Checkmk dashboard" doesn't update anymore after version update

CMK version: 2.0.0p20 cre
OS version: Oracle Linux 8.5

Error message: nil.

Hi all,
We installed a new infrastructrure with distributed monitoring. We started it on version 2.0.0 with some differente p levels so, before go live, we updated every site to the same 2.0.0p20 (so we had also tested the update procedure).
Every site was correctly updated and it was up and running but the “service overview” reported in the checkmk dashboard didn’t updates anymore for some sites (not for all).
In those sites while updating its services went down as expected but never come back to “running” state even when the site had become correctly running again.
So I searched a bit around on the master site filesystem and I found a folder that appear to be cache and I remove it. It was not a great idea, it didn’t solve and left the dashboard in this state:

Any idea on how to resolve?

The information on that page is mostly asynchronous.
First restore whatever you deleted.
Then check if the information appears after some time.
As said, what you see there is not real-time.

Well, I can’t restore it: It appear to was cache so I didn’t backup it.
I waited at least 10 day before try to reset the dashboard cache so I think was not due to impatience!

Any idea about how to force rebuilding?
Disconnect/Remove a site from distributed monitoring and add it again which kind of impact could have?

Can you create a new site and check whether the dashboard works there when adding the hosts?

Yes, after adding a new site it was correctly added to the dashboard in the upper part (Checkmk servers) but not in bottom part (“Site overview”) that it is not updated.

Hi @frakka is it solved?

No, the issue still persist.
I hope it magically solves upgrading to 2.1.x but I cannot do it yet.

In your distributed RAW environment you need to sync the HW/SW inventory data between your instances and the master site.
In a Enterprise environment this sync should happen automatically.

I’ll verify immediatly when back to work.
I don’t remenber nothing about this requirement in the docs…
Thank you for your suggestion.

Are you talking about this?

I’ll do, thank you.

I just checked and on the master site I have the directory “~/var/check_mk/inventory” and it contains a UTF-8 Unicode text and a gzip compressed data archive for each site in my distributed environment.
They seems to be update (modification date are just few minutes ago and they are all coordinated between plain text files and gz archive).
I haven’t a sync jobs yet so those files are not updated by an rsync process or similar…

Instead, on the distributed monitored sites the directory “~/var/check_mk/inventory” doesn’t exists, nor in satellites which appear in the checkmk dashboard nor on satellites which doesn’t appear.

The plugin file “/usr/lib/check_mk_agent/plugins/mk_inventory.linux” and its configuration “/etc/check_mk/mk_inventory.cfg” exists and the plugin is executable on all sites.
I left the “INVENTORY_INTERVALL” as default.

What I’m doing wrong?

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed. Contact an admin if you think this should be re-opened.