Www.wolters-cat-dog.de crashes kresd

Hi,

I have a TurrisOS 4.0.5 on a Mox and see the problem that the above name seems to crash kresd, i.e. it is not showing in “ps aux” anymore.
See below for the steps to reproduce the problem. I had to replace the real name to satisfy the posting policy :frowning:
Any idea what is so special about this name and how I could debug the problem?

Peter

P.S.: Many thanks for making Mox. It is an awsome device!

root@turris:~# nslookup XXXXX
Server: 127.0.0.1
Address: 127.0.0.1#53

Name: XXXXX
Address 1: 216.58.210.4
Address 2: 2a00:1450:4001:821::2004
root@turris:~# nslookup www.wolters-cat-dog.de
;; connection timed out; no servers could be reached

root@turris:~# nslookup XXXXX
;; connection timed out; no servers could be reached

root@turris:~# /etc/init.d/kresd restart
syntax error. Last token seen: +
Garbled time
root@turris:~# nslookup XXXXX
Server: 127.0.0.1
Address: 127.0.0.1#53

Name: XXXXX
Address 1: 172.217.18.4
Address 2: 2a00:1450:4001:80b::2004

:thinking: the issue doesn’t happen for me on current 3.11.x and HBT (:turtle:5.0.0) and my PC using kresd version from Turris 4.0.5. If you don’t mind:

  • Is it repeatable for you?
  • Is there something in logs from the moment it crashes?

Yes, it is repeatable.

    root@turris:~# nslookup www.wolters-cat-dog.de
;; connection timed out; no servers could be reached

root@turris:~# nslookup www.google.de
Server:		127.0.0.1
Address:	127.0.0.1#53

Name:      www.google.de
Address 1: 172.217.23.163
Address 2: 2a00:1450:4001:81f::2003
root@turris:~# nslookup www.google.de
;; connection timed out; no servers could be reached

root@turris:~# nslookup www.google.de
;; connection timed out; no servers could be reached

root@turris:~# ps aux | grep kresd
root      4954  0.0  0.0   1248   608 pts/0    S+   16:21   0:00 grep kresd


root@turris:/etc/config# logread -f
May  3 16:12:02 turris syslog-ng[4349]: syslog-ng starting up; version='3.21.1'
May  3 14:14:57 turris kresd[4504]: [ ta ] key: 20326 state: Valid
May  3 14:14:57 turris kresd[4504]: [ ta ] next refresh for . in 7.5066666666667 hours
May  3 14:15:01 turris /usr/sbin/cron[4521]: (root) CMD (/usr/bin/notifier)
May  3 14:20:01 turris /usr/sbin/cron[4849]: (root) CMD (/usr/bin/notifier)
May  3 14:20:03 turris kresd[4834]: [ ta ] key: 20326 state: Valid
May  3 14:20:03 turris kresd[4834]: [ ta ] next refresh for . in 7.5066666666667 hours
May  3 14:20:10 turris procd: Instance kresd::instance1 s in a crash loop 6 crashes, 17 seconds since last crash
1 Like

Thank you. On second thought… I think deeper investigation isn’t worth it unless it gets significantly worse or gets reproduced on recent Turris OS 3.x or 5.x. I think 4.x does not plan to go beyond 4.0.5 anyway, and due to some technical packaging problems it contains quite an old version of kresd.

In case you were interested in “early” update to 5.x, here’s the thread: Turris OS 5.0.0 is released in HBT – but I hope that even the slowest >= 4.x channel will switch to 5.x soon.

You are probably right. I would say the same if I were in your place :slightly_smiling_face:

I was looking at 5.0.0 but due to Covid I rely very much on a stable network at home. So I didn’t dare to update to a beta. Is there already a date for the stable release in sight?

Peter

The best I know is to watch that thread. Multiple people (including myself) apparently use 5.0.0 without any issue, but some did report issues, as you can see there. I believe rollbacks should work the same between 4.x and 5.x…

It was mentioned here

It would suggest that a release of TOS5.x is imminent now that OpenWrt 18.06 has reached EoL.

1 Like