Question #12844
closedsome servers grey
Description
Hello,
I have started deploying rudder ESR (4.1) in our company and so far everything worked well.
Since the last update (4.1.12) I have noticed about 10 nodes (out of 176) that are grey. Some are behind
relays, others connect directly to the server.
Last seens is always old for these servers. When checking the agent with either rudder agent health/status/check
no errors are reported.
I can run rudder agent run/inventory but state in dashboard doesn't change.
On some servers I have already removed rudder, deleted all remaining files and reinstalled but it didn't help.
Could anybody please help me with debugging this behaviour? I am not sure what could be causing this.
Updated by Alexis Mousset almost 6 years ago
Hello Tobias,
What is the message displayed in the details of one of these nodes, in the "Compliance" tab? It could help understand the problem. Furthermore:
- Reports are sent to the server using syslog, is the syslog service (probably rsyslog or syslog-ng) correctly running on these nodes?
- Does "rudder agent update" work on these nodes? If not, it could also explain the missing compliance.
Updated by Tobias Ell almost 6 years ago
Hello Alexis,
thanks for helping with this problem.
Message in compliance is:
"No recent reports have been received for this node in the grace period since the last configuration policy change. This is unexpected. Please check the status of the agent by running 'rudder agent health' on the node.
For information, expected node policies are displayed below.
Current configuration ID for this node is '20180626-105840-afddbf11' (generated on 2018-06-26 10:58:40)."
Running "rudder agent helath" simply says "OK".
The local rsyslogd is running and "rudder agent update" says "ok: Rudder agent promises were updated."
Updated by Tobias Ell almost 6 years ago
Updated by Benoît PECCATTE over 5 years ago
- Related to Bug #12870: rudder server not processing some reports added
Updated by Benoît PECCATTE over 5 years ago
Sorry for the delay.
The related issue has been closed, so i'm closing this question.
Feel free to reopen if there is still a problem.