Timeout at automatic update

We try to enable automatic agent update in our multisite configuration and it’s working pretty fine except for one site which has a rather low-speed connection.

Registration went fine but when doing the update on this site we always run into timeouts and the updater reports:

Agent Bakery: Error communicating with central bakery: Your request timed out after 110 seconds. This issue may be related to a local configuration problem or a request which works with a too large number of objects. But if you think this issue is a bug, please send a crash report.

Is there any possibility to change this timeout?

Thanks!

@MasopustC did you ever get this resolved? We implemented SSL and now the agents in one site can’t connect to the repository despite baking the agents with the correct FQDN of the local CheckMK server and https connection parameters.

Any thoughts?

@scott.reiser
First thought: Did you check that you provide all neccessary CA certificates for the SSL connection? If there are different CA chain in use you must provide all of them with the agent update plugin and make sure that they are distributed to all clients before those CA chains are used.
(This is not neccessary if CA chains are used with Root CAs that are included in the official CA stores, but if more CA chains are used, these also have to be configured explicitly for the agent update according to the handbook:
" Important: If the server’s certificate chain is signed with a public CA, the connection can normally be verified without imported certificates. However, as soon as imported certificates from one of the mentioned sources are available to the agent updater, all other CA certificate authorities are ignored! In the case of problems with the configuration of HTTPS, please consult the following FAQ below.")

Is it just one site that doesn’t work correctly while other sites work fine? What does the /var/lib/check_mk_agent/cmk-update-agent.log of one of the failed agents say? Are there any error messages that might help solve the problem?

Thank you for your reply @tawi. Turns out we had a couple of other environmental issues going on related to the remote, standalone CheckMK instance (and deployed agents). What appeared to be CA and SSL cert mismatching was actually problems with the agent bakery and agent auto update settings at the bakery as well as at the agent level (cmk-update-agent.cfg).

Pulling the remote sites into our CheckMK core site, left agent configs that pointed to the old bakeries.

Fortunately, the remote sites were small and mostly for dev so I blew them away and set them up from scratch, baked and reinstalled new agents.

Sorry for late reply. We simply solved it by increasing the speed of the connection to the remote site :slight_smile:

As we have now v2.1, is there any way how to increase hardocded 110 seconds timeout?

Hi! I am not sure if this is the same issue you asked about, but there was a similar question recently: Can web ui 110s timeout be incresed? – it has a seemingly useful reference.

Maybe this could help in your case as well?

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.