Is this normal?

Hi,

Browsing through the system log; I see these warnings appear every 10 minutes. Is this normal?

2016-12-18T03:25:03+01:00 warning watchdog[]: Restarted resolver
2016-12-18T03:25:03+01:00 warning watchdog[]: Restarted nethist

I have also “warning watchdog[]: Restarted nethist” messages but no the other ones.

similar here:

[ root@turris ] ~
$> cat /var/log/messages  | grep watchd | tail -n 20
2016-12-18T15:15:02+01:00 warning watchdog[]: Restarted nethist
2016-12-18T15:20:01+01:00 info /usr/sbin/cron[28074]: (root) CMD (/usr/bin/watchdog.sh)
2016-12-18T15:25:01+01:00 info /usr/sbin/cron[31955]: (root) CMD (/usr/bin/watchdog.sh)
2016-12-18T15:25:02+01:00 warning watchdog[]: Restarted nethist
2016-12-18T15:30:01+01:00 info /usr/sbin/cron[3171]: (root) CMD (/usr/bin/watchdog.sh)
2016-12-18T15:35:01+01:00 info /usr/sbin/cron[6285]: (root) CMD (/usr/bin/watchdog.sh)
2016-12-18T15:35:02+01:00 warning watchdog[]: Restarted nethist
2016-12-18T15:40:01+01:00 info /usr/sbin/cron[8724]: (root) CMD (/usr/bin/watchdog.sh)
2016-12-18T15:45:01+01:00 info /usr/sbin/cron[11323]: (root) CMD (/usr/bin/watchdog.sh)
2016-12-18T15:45:02+01:00 warning watchdog[]: Restarted nethist
2016-12-18T15:50:01+01:00 info /usr/sbin/cron[13865]: (root) CMD (/usr/bin/watchdog.sh)
2016-12-18T15:55:01+01:00 info /usr/sbin/cron[14143]: (root) CMD (/usr/bin/watchdog.sh)
2016-12-18T15:55:02+01:00 warning watchdog[]: Restarted nethist
2016-12-18T16:00:01+01:00 info /usr/sbin/cron[14439]: (root) CMD (/usr/bin/watchdog.sh)
2016-12-18T16:05:01+01:00 info /usr/sbin/cron[15452]: (root) CMD (/usr/bin/watchdog.sh)
2016-12-18T16:05:02+01:00 warning watchdog[]: Restarted nethist
2016-12-18T16:10:01+01:00 info /usr/sbin/cron[18998]: (root) CMD (/usr/bin/watchdog.sh)
2016-12-18T16:15:01+01:00 info /usr/sbin/cron[21744]: (root) CMD (/usr/bin/watchdog.sh)
2016-12-18T16:15:02+01:00 warning watchdog[]: Restarted nethist
2016-12-18T16:20:01+01:00 info /usr/sbin/cron[25365]: (root) CMD (/usr/bin/watchdog.sh)
[ root@turris ] ~
$>

OK, so “warning watchdog[]: Restarted nethist” seems normal.
I disabled resolver (with kresd) because of my issue with slow opening of websites. It is logged that resolver is restarted. If i look under Status-processes or System-software then resolver is not present c.q. is disabled. Why than log it that it is restarted?

I am seeing watchdog warnings about both resolver and nethist being restarted. These occur together, and right about 10min apart as you were observing.

I’ve started to see kresd errors like “bind to ‘0.0.0.0#53’ address is in use”; and “worker failed. No such file or directory.” Although dnsmasq appears to be operating based on log messages (it is servicing a few devices), my dhcp-assigned servers have once again dropped off network. And I am periodically seeing debug kernel messages related to ath10k_pci ATH10K_DBG_BUFFER.

That means you have something already listening port 53. You can try command “netstat -anp | grep -w :53” to find out what is it.

Hi @white,

Here’s the result from netstat -anp | grep -w :53
tcp 0 0 :::53 :::* LISTEN 5753/kresd
udp 0 0 :::53 :::* 5753/kresd

I scrolled back in the tail -f output of messages without finding any “address in use” messages recently. I also found that kresd, while running, has become disabled in the startup page. Kresd had been enabled when I last viewed the startup page checking on dnsmasq. I’m wondering if programs are kicked into disabled based on errors? That could explain why dnsmasq was disabled yesterday.

Additionally for the last several hours, I’ve once again had no functioning dhcp. Here’s the dnsmasq process info:
2121 nobody 876 S /usr/sbin/dnsmasq -C /var/etc/dnsmasq.conf -k -x /var/run/dnsmasq/dnsmasq.pid

I saw a thread about “dnsmasq running as nobody” on the forum - I’ll check that out next.

This has been a very bizarre several days battling dhcp issues. My network runs, then fails. Rebooting to last snapshot config recovered my Omnia earlier today; but has done nothing to remediate dhcp failures this afternoon. Maybe a new stapshot replaced the old. I’ll look into that also.

Can I somehow get rid of
warning watchdog[]: Restarted nethist
?

That service shutdown for some reason? If watchdog must restart it all the time?

Also those
CMD (/usr/bin/rainbow_button_sync.sh) CMD (nethist_stats.lua)

are annoying, as I dont have LEDs actual visible :slight_smile: But at least, its just cron info, but warning above really interest me…

Does it still happen for resolver? If you have connection problems, the watchdog should try restarting the DNS resolver. It checks for ping turris.cz working: https://gitlab.labs.nic.cz/turris/misc/blob/master/watchdog/watchdog.sh#L51

If you mean nethist restarts, yes… I dont think it has something with network connection.

Watchdog didnt restart anything else, ¨just nethist, every 10mins…

OK. I meant Restarted resolver, as in the first post.