Hi all,
i have an auto-update in my distributed env and i have some Redhat agents which are using the wrong url for contacting the master for updates.
I defined an url for my master in the updater rule, but the agent is trying to contact the slave site to which it is connected first. The slave site cannot be contacted from the agent and so the update is waiting for a 2 minutes timeout before trying the correct master url.
I also defined the “always” option in the rule and as far as i understand the updater should always use the defined url for updates.
But it does not. So my update takes always about 6 minutes to finish.
Maybe in the first registration command someone manually configured the wrong satellite or the config in the Installation packages was wrong and the hosts were registered to the satellite by accident in the first place.
If the config file is correct, you can delete the state file and reregister the host again.
Hi Andre, i am so sorry, but i lost this topic last year.
I checked the state file and it includes the remote site url instead of the master site. I modified the url manually and started the updater. It contacts the master as wanted but than updated the url again to the remote site automatically.
I am not sure where this info is comming from. I added the master url in global settings too, but it always changed back.
the configuration is done in the checkmk agent bakery rules.
You can create duplicate entries for the rule “Agent updater (Linux, Windows, Solaris)”, where you then assign the correct sattelite based on the folder structure.
