/var/log getting full

[calypsoSA]

Hello

Just had this error come up. I've taken a noob look but cant quite see why this is happening. Moving logging is off, server been on 30 days, 10 or so docker containers?

 

Can anyone tell me where to hunt the issue in the diagnostics.

 

Thanks

tower-diagnostics-20180617-1203.zip

[calypsoSA]

Any pointers of where I can look or what to look out for?

[pwm]

Just curious - did you take a quick look in the logs yourself?

Jun 11 05:51:39 Tower shfs: cache disk full
Jun 11 05:52:00 Tower shfs: cache disk full
Jun 11 05:52:08 Tower shfs: cache disk full
Jun 11 05:52:08 Tower shfs: cache disk full
Jun 11 05:52:09 Tower shfs: cache disk full
Jun 11 05:52:09 Tower shfs: cache disk full
Jun 11 05:52:10 Tower shfs: cache disk full
Jun 11 05:52:10 Tower shfs: cache disk full
Jun 11 05:52:12 Tower shfs: cache disk full
Jun 11 05:52:12 Tower shfs: cache disk full
Jun 11 05:52:30 Tower shfs: cache disk full
Jun 11 05:52:30 Tower shfs: cache disk full
Jun 11 05:52:33 Tower shfs: cache disk full
Jun 11 05:52:41 Tower shfs: cache disk full
Jun 11 05:52:41 Tower shfs: cache disk full
Jun 11 05:53:00 Tower shfs: cache disk full
Jun 11 05:53:08 Tower shfs: cache disk full
Jun 11 05:53:08 Tower shfs: cache disk full
Jun 11 05:53:11 Tower shfs: cache disk full
Jun 11 05:53:11 Tower shfs: cache disk full
Jun 11 05:53:12 Tower shfs: cache disk full
Jun 11 05:53:12 Tower shfs: cache disk full
Jun 11 05:53:13 Tower shfs: cache disk full
Jun 11 05:53:13 Tower shfs: cache disk full
Jun 11 05:53:30 Tower shfs: cache disk full
Jun 11 05:53:30 Tower shfs: cache disk full

And I think you have a broken docker image - probably as a result of having a full cache volume.

[calypsoSA]

I did and a fixed the error from the cache getting full. 

I just had no idea that the simple file could get so big that it becomes a capacity issue. Now I know. 

 

Thanks for the help. 

[pwm]

15 minutes ago, calypsoSA said:

I did and a fixed the error from the cache getting full. 

I just had no idea that the simple file could get so big that it becomes a capacity issue. Now I know. 

 

Thanks for the help. 

 

That's the compromise you get from running the full system in RAM. It would be nice if the system could send out a mail if the log seems to be spammed by specific log lines instead of having to wait until the system notices that the log file has already become huge.

 

Another thing I hope unRAID could look into is supporting forwarding of log data to another machine. The log daemons in Linux have had this support for many, many years. On one hand, it allows a single focal point for the logs of many machines with helps with system supervision. On the other hand, it means that a machine that gets hacked has already sent out the log data so the hacker can't rewrite the log data to remove incriminating log lines.