+ dan for input, also marking as a fixit candidate

We have disk space monitoring in all lab servers through nagios. For devservers in ACL enabled subnet, this data is not collected.

What's the alert's keyword? 

I searched in inbox, and can't find the alerts of chromeos-server74.cbf (the "down" shard on Aug 16th's morning.

https://groups.google.com/a/google.com/forum/#!searchin/chromeos-build-alerts/chromeos-server74$20disk%7Csort:relevance/chromeos-build-alerts/kZTPN9UZrLI/8OEFuFTqBgAJ

Root cause of that shard is some kind of disk IO issue, which led to gs_offloader taking long time to upload results.

That shard is full at Aug 16 morning, makes many service not work properly. This won't trigger an alert at that time?

I see the newest alert of that shard that reporting root partition problem is on Wed Aug 17 12:17:24, not the same day.

some sad background info. The shard was added to hosts.cfg in nagios by Charlene a while ago, but we didn't restart nagios, that's why the server was never monitored...

oh! understood~ 

> some sad background info. The shard was added to hosts.cfg
> in nagios by Charlene a while ago, but we didn't restart
> nagios, that's why the server was never monitored...

When human error occurs, it's a bug in the software.

Can someone file a _new_ bug requesting that we create foolproof
administrative mechanisms so that this can't happen?

When human error occurs, it's a bug in the software.

－－－ is it a slogan from some books?


Bug filed: https://bugs.chromium.org/p/chromium/issues/detail?id=640442