The diskspace monitoring is broken and not taking offline/online agents by itself in the background. It only happens when one visits the .../computer page.

Steps to reproduce

1. start Jenkins with -Dhudson.node_monitors.AbstractNodeMonitorDescriptor=3 (so that the monitors run every 3 minutes)
2. add an agent
3. configure free/temp disk space monitor to a value that is slightly above the current free space
4. verify agent is online
5. create a file on the agent so that the free space gets below threshold
6. wait 3 minutes, do not visit the .../computer page during that time

Similarly when the agent is already offline, it is not taken online again unless one visits .../computer

Expected Result

The agents is taken offline with a message about not enough disk space

Actual Result

The agent is still online

When one now visits .../computer, the agent is taken offline immediately. Another way to trigger that an agent is taken offline/online is to visit the api endpoint of an agent