Syslog full

[Zodis]

My syslog keeps getting full, it's set to the default size in unraid. Lately I've been using the command mount -t tmpfs -o remount,size=50% tmpfs /var/log to keep it at a maximum of 50% of available ram, but this is only a temporary fix since after a system reboot it reverts back to the default size. I'd like to know how to permanently expand the syslog size and I've also attached a diagnostic of my server if anyone could tell me what's causing the problem. 

thebesttower-diagnostics-20171010-2042.zip

[itimpi]

Looks like your syslog is getting spammed by error messages relating to a (I assume non-existent) floppy disk drive.    Have you made sure that floppy disk support is disabled in your BIOS settings?

 

if you want to make a change to the syslog size persist across reboots then you need to add the command required to the ‘go’ file in the config folder on the flash drive.

[Zodis]

Thanks for the info I'll take a look tonight.

 

So if the go file says 

 

#!/bin/bash
# Start the Management Utility
/usr/local/sbin/emhttp &

mount -t tmpfs -o remount,size=50% tmpfs /var/log

 

Then that should work?

[itimpi]

6 hours ago, Zodis said:

Thanks for the info I'll take a look tonight.

 

So if the go file says 

 

#!/bin/bash
# Start the Management Utility
/usr/local/sbin/emhttp &

mount -t tmpfs -o remount,size=50% tmpfs /var/log

 

Then that should work?

It probably will, but I wonder if it is better to do the resize before starting emhttp?

[Zodis]

Thanks a lot for the advice. I disabled the floppy controller in the BIOS and now the syslog doesn't have that error showing up constantly. I also set the go file to be 

 

#!/bin/bash
# Start the Management Utility
mount -t tmpfs -o remount,size=50% tmpfs /var/log
/usr/local/sbin/emhttp &

 

[nt7561]

Hi everyone. I also have an issue with my syslog. I'm attaching it. Can anyone please advice what can I do?

Many thanks in advance.

tower-diagnostics-20220609-1850.zip

[JorgeB]

9 minutes ago, nt7561 said:

I also have an issue with my syslog.

 

This is the repeating line:

 

Jun  8 02:56:31 Tower kernel: w83795 0-002f: Failed to read from register 0x02f, err -6

 

Can you try booting in safe mode? Or since it looks sensor related try uninstalling the system temp plugin.

[nt7561]

31 minutes ago, JorgeB said:

 

This is the repeating line:

 

Jun  8 02:56:31 Tower kernel: w83795 0-002f: Failed to read from register 0x02f, err -6

 

Can you try booting in safe mode? Or since it looks sensor related try uninstalling the system temp plugin.

OK I will uninstall the plug in and check it out. Thanks again for the tip 

[nt7561]

1 hour ago, JorgeB said:

 

This is the repeating line:

 

Jun  8 02:56:31 Tower kernel: w83795 0-002f: Failed to read from register 0x02f, err -6

 

Can you try booting in safe mode? Or since it looks sensor related try uninstalling the system temp plugin.

I think problem solved. I've uninstalled the system temp plugin rebooted and now the error disappeared. 

[gstacks13]

I'm also having the issue of a full syslog.  Mind helping me find the culprit?

sheik-diagnostics-20221130-1342.zip

[Squid]

Run check disk filesystem against disk 2  https://wiki.unraid.net/Check_Disk_Filesystems

[gstacks13]

You nailed the issue. I had just swapped in a new Disk 2, and Check Disk Filesystems found a handful of errors after the data rebuild.  Running the filesystem repair and rebooting solved the issue.  In hindsight I should've formatted the drive before starting the data rebuild, but we still got there in the end!

 

Thanks for your help!

[JorgeB]

13 hours ago, gstacks13 said:

In hindsight I should've formatted the drive before starting the data rebuild

Only if there wasn't any data there, or everything would be lost.

[gstacks13]

Yep, and there wasn't. I was swapping in a brand new drive to replace a failing one. So formatting would've been fine in that case.

 

But, you live and you learn!

[JorgeB]

OK, just wanted to make sure, some users think they can format a drive and restore the data with a disk rebuild.

[taylorjdunn]

Hello, I've run into a similar issue. syslog shows as 200K, which I assume is too big. I also noticed that Tailscale's log file is about 160K.

 

I've confirmed that logging is disabled in the GUI, and a restart seems to fix it temporarily. Any recommendations?

yosemite-diagnostics-20240208-2032.zip

[trurl]

10 minutes ago, taylorjdunn said:

syslog shows as 200K, which I assume is too big. I also noticed that Tailscale's log file is about 160K.

Neither of those are too big, log space is 128M. But log space is getting filled by something.

 

What do you get from command line with this?

du -h -d 1 /var/log

 

[taylorjdunn]

13 hours ago, trurl said:

Neither of those are too big, log space is 128M. But log space is getting filled by something.

 

What do you get from command line with this?

du -h -d 1 /var/log

 

Oh OK, must be something else then. Here's the results from that:

 

0       /var/log/pwfail
126M    /var/log/unraid-api
0       /var/log/swtpm
1.7M    /var/log/samba
0       /var/log/plugins
0       /var/log/pkgtools
4.0K    /var/log/nginx
0       /var/log/nfsd
0       /var/log/libvirt
128M    /var/log

 

So log is the full 128M, and looks like the majority is contained within /unraid-api

[itimpi]

46 minutes ago, taylorjdunn said:

So log is the full 128M, and looks like the majority is contained within /unraid-api

Are you on the latest release of the Connect plugin and have you rebooted to release the current log space since going to the latest release of that plugin?  An earlier release had a bug that was doing excessive logging and could cause these symptoms

[taylorjdunn]

10 minutes ago, itimpi said:

Are you on the latest release of the Connect plugin and have you rebooted to release the current log space since going to the latest release of that plugin?  An earlier release had a bug that was doing excessive logging and could cause these symptoms

Just updated to the current version and rebooted. Hopefully that's the fix, thanks for your help! I'll reply back if it keeps happening.