update interval > 60 causes server to be marked as inactive

[chris-y]

Problem description

If the update_interval in the sidecar.yml file is set to a value higher than 60, Graylog marks the server as inactive after 60 seconds until the next update.

Graylog should receive the information about how often the server/sidecar sends updates and base the "inactive" flag on that.

Steps to reproduce the problem

1. Set update_interval: 300
2. Wait about a minute
3. Server shows as inactive in the Sidecar area in Graylog until the next update

Environment

• Sidecar Version: 1.2.0
• Graylog Version: 4.3.7
• Operating System: Windows
• Elasticsearch Version: N/A
• MongoDB Version: N/A

[thll]

Hi @chris-y, you can globally configure the threshold that Graylog uses to determine if a sidecar is considered inactive. You can find the setting on the System/Configurations page:

If you raise that high enough to cover your largest update interval, it should fix the issue.

This is not a perfect solution if you configure your sidecars to have different update intervals, but in most cases, it should be good enough.

[chris-y]

Hi, yes that has worked, thank you. Is this documented anywhere? As it's a bit hidden away!

[thll]

I think the docs are indeed missing. I've added it to our list of documentation TODO items.