I woke up this morning to login to the webUI and see that all my disks are gone. I tried to go into tools to run diagnostics but it hangs when I click on the tools tab. All dockers seem to work fine and I can even watch videos on plex, and I can SSH into my server. Still have no idea the cause of this problem and haven't restarted my server yet. I've also tried tried to type diagnostics in ssh but it's giving me an error that there is not enough free diskspace. I've tried multiple browsers, cleared cache and it still shows the same results.

 

 

DF command gives:

 

Diagnostics command gives:

 

 

Edited July 25, 20178 yr by bobokun

Before you reboot, do this from the command line

 

cp /var/log/syslog /boot/syslog.txt

 

and then post the file.

Something has filled your RAM memory, it says your rootfs (16GB) is 100% used.

 

Usual culprits are wrong path definitions for dockers.

17 minutes ago, Squid said:

Before you reboot, do this from the command line

 

 

cp /var/log/syslog /boot/syslog.txt

 

 

and then post the file.

 

syslog.txt

10 minutes ago, bonienl said:

Something has filled your RAM memory, it says your rootfs (16GB) is 100% used.

 

Usual culprits are wrong path definitions for dockers.

Not sure why it's only 16GB...I have 32GB of RAM installed 

3 minutes ago, bobokun said:

Not sure why it's only 16GB...I have 32GB of RAM installed 

 

It is 16GB for rootfs, the remaining 16GB is for the other filesystems.

 

Your syslog has many warnings like this "... failed parsing crontab for user root: docker exec $DOCKER_NAME pihole..."

At some point this is causing out of memory issues and consequently all kinds of errors. Did you create some cronjob?

Yes I got the cronjob script based on what was recommended by @spants in the pihole docker support thread:

 

Edited July 25, 20178 yr by bobokun

12 minutes ago, bobokun said:

Yes I got the cronjob script based on what was recommended by @spants in the pihole docker support thread:

 

I think its safe to say that you did it wrong.  You should post what you actually did.

 

Curiously, you also have "uninitialized csrf_token" errors, and I've been wondering how its possible to actually get that error, and perhaps full rootfs is one way.

 

1 minute ago, Squid said:

I think its safe to say that you did it wrong.  You should post what you actually did.

 

Curiously, you also have "uninitialized csrf_token" errors, and I've been wondering how its possible to actually get that error, and perhaps full rootfs is one way.

 

So there was two ways I did this. I probably only need to do one. The first method I did this was create a file called pihole.cron in the /boot/config/plugins/pihole directory. The second method was to do it through userscripts. Both the scripts in user.scripts plugin and the one pihole.cron have the same script.

 

 

 

53 minutes ago, bobokun said:

The first method I did this was create a file called pihole.cron in

That's the problem.

 

Your pihole.cron file should be something like

1 2 * * *  /boot/myscripts/piholescript

 

Where the 1 2 * * * is a cron expression for whenever to run, followed by a space followed by a path to the script to actually run.

 

Saving the script itself as a .cron file winds up having cron trying to parse the file which won't work.

 

To be honest though, I'm not quite sure how that would result in rootfs getting full (I've never tried making the mistake on purpose), and my gut is still @bonienl suggestion of 

1 hour ago, bonienl said:

Usual culprits are wrong path definitions for dockers.

But, fix the cron thing first so at least the syslog gets cleaned up.

you are going to have to reboot the server though.  

Edited July 26, 20178 yr by Squid

1 minute ago, Squid said:

That's the problem.

 

Your pihole.cron file should be something like

 

1 2 * * *  /boot/myscripts/piholescript[code]

 

Where the 1 2 * * * is a cron expression for whenever to run, followed by a space followed by a path to the script to actually run.

 

Saving the script itself as a .cron file winds up having cron trying to parse the file which won't work.

 

To be honest though, I'm not quite sure how that would result in rootfs getting full (I've never tried making the mistake on purpose), and my gut is still @bonienl suggestion of 

But, fix the cron thing first so at least the syslog gets cleaned up.

you are going to have to reboot the server though.  

Thanks for the explanation. I'm still new to using unraid so I didn't realize that the output of pihole.cron is not the script itself. I'll use method 2 instead through userscripts and delete the pihole.cron file. That should fix the cron issue

Did you start one or more preclear actions?

 

6 hours ago, bonienl said:

Did you start one or more preclear actions?

 

yes I was in the process of preclearing two drives