Strange behaviour of bash script in crontab
I have a crontab entry which runs ones per hour calling a bash script. The content of the script is: #! /bin/bash # # dig smtp.gmail.com >> digoutput.txt 2>&1 rc=$? [ $rc -ne 0 ] && echo "exitcode dig = $rc" The output of the cronjob is: /home/xxxxx/bin/script.sh: line 4: 32349 Terminated dig smtp.gmail.com >> digoutput.txt 2>&1 exitcode dig =143 143 means the dig command is terminated by SIGTERM Any idea? -- fr.gr. member openSUSE Freek de Kruijf
Op zaterdag 25 juni 2022 14:37:21 CEST schreef Dave Howorth:
On Sat, 25 Jun 2022 11:54:21 +0200 Freek de Kruijf <freek@opensuse.org> wrote:
dig smtp.gmail.com >> digoutput.txt 2>&1
what happens if you run this command by itself in a terminal?
It works as expected, but this only happens occasionally, normally the command works OK. Yesterday it occurred 4 times out of 24 and today it already occurred 4 times out of 16. It may be important to mention: this is with Leap 15.4 on a Raspberry Pi 4B. -- fr.gr. member openSUSE Freek de Kruijf
Remember that with cron you need to set the path env as there is no place it is inherited from. Ken Schneider
On Jun 25, 2022, at 9:25 AM, Freek de Kruijf <freek@opensuse.org> wrote:
Op zaterdag 25 juni 2022 14:37:21 CEST schreef Dave Howorth:
On Sat, 25 Jun 2022 11:54:21 +0200 Freek de Kruijf <freek@opensuse.org> wrote:
dig smtp.gmail.com >> digoutput.txt 2>&1
what happens if you run this command by itself in a terminal?
It works as expected, but this only happens occasionally, normally the command works OK. Yesterday it occurred 4 times out of 24 and today it already occurred 4 times out of 16.
It may be important to mention: this is with Leap 15.4 on a Raspberry Pi 4B.
-- fr.gr.
member openSUSE Freek de Kruijf
Freek, et al -- ...and then Freek de Kruijf said... % % It works as expected, but this only happens occasionally, normally the % command works OK. Yesterday it occurred 4 times out of 24 and today it % already occurred 4 times out of 16. That's very odd indeed. It doesn't sound like it's related to runnimg under cron but instead something transient in the system. I was going to suggeat that you strace dig, but i don't think that;s worth it. % % It may be important to mention: this is with Leap 15.4 on a Raspberry Pi 4B. I would lean next toward a resource limitation, and although a Pi is awesome :-) perhaps it's struggling. What does top tell you? What's the uptime? Carlos usually has great ideas here :-) You might try adding some debug info into the script to give you a picture of the world each time cron kicks it off. What if you run the command manually oh, say, 1000 times with no waiting? :-D -- David T-G See http://justpickone.org/davidtg/email/ See http://justpickone.org/davidtg/tofu.txt
Op maandag 27 juni 2022 12:32:03 CEST schreef David T-G:
Freek, et al --
...and then Freek de Kruijf said... % % It works as expected, but this only happens occasionally, normally the % command works OK. Yesterday it occurred 4 times out of 24 and today it % already occurred 4 times out of 16.
That's very odd indeed. It doesn't sound like it's related to runnimg under cron but instead something transient in the system. I was going to suggeat that you strace dig, but i don't think that;s worth it.
% % It may be important to mention: this is with Leap 15.4 on a Raspberry Pi 4B.
I would lean next toward a resource limitation, and although a Pi is awesome :-) perhaps it's struggling. What does top tell you? What's the uptime? Carlos usually has great ideas here :-) You might try adding some debug info into the script to give you a picture of the world each time cron kicks it off. What if you run the command manually oh, say, 1000 times with no waiting?
:-D
Tried it 1000 times, without waiting and a sleep 1. Both reported one issue, respectively on count 450 and 179. top i shows: top - 16:44:13 up 13 days, 2:02, 1 user, load average: 0.01, 0.05, 0.02 Tasks: 180 total, 1 running, 179 sleeping, 0 stopped, 0 zombie %Cpu(s): 0.2 us, 0.2 sy, 0.0 ni, 99.6 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st MiB Mem : 7836.246 total, 1192.496 free, 1704.031 used, 4939.719 buff/cache MiB Swap: 8195.996 total, 8194.246 free, 1.750 used. 6051.578 avail Mem -- fr.gr. member openSUSE Freek de Kruijf
Freek, et al -- ...and then Freek de Kruijf said... % % Op maandag 27 juni 2022 12:32:03 CEST schreef David T-G: % > ... % > uptime? Carlos usually has great ideas here :-) You might try adding % > some debug info into the script to give you a picture of the world each % > time cron kicks it off. What if you run the command manually oh, say, % > 1000 times with no waiting? % > % Tried it 1000 times, without waiting and a sleep 1. Both reported one issue, % respectively on count 450 and 179. [snip] Oho! At least you can rule out cron, which makes things one step simpler. Yay. Time to turn this back over to the experts. I'm just an out-of-work admin who fumbles around OK :-) But a big branch of the decision tree has just been pruned away, so I can feel like I helped. :-D -- David T-G See http://justpickone.org/davidtg/email/ See http://justpickone.org/davidtg/tofu.txt
Hi there, On Mon, 27 Jun 2022, 16:51:55 +0200, David T-G wrote:
Freek, et al --
...and then Freek de Kruijf said... % % Op maandag 27 juni 2022 12:32:03 CEST schreef David T-G: % > ... % > uptime? Carlos usually has great ideas here :-) You might try adding % > some debug info into the script to give you a picture of the world each % > time cron kicks it off. What if you run the command manually oh, say, % > 1000 times with no waiting? % > % Tried it 1000 times, without waiting and a sleep 1. Both reported one issue, % respectively on count 450 and 179. [snip]
Oho! At least you can rule out cron, which makes things one step simpler. Yay.
Time to turn this back over to the experts. I'm just an out-of-work admin who fumbles around OK :-) But a big branch of the decision tree has just been pruned away, so I can feel like I helped.
this smells like what I have seen every once in a while when resolving host names with an IPv6 address. At least I don't have IPv6 available here, it ended up with error messages; this mostly happened with ntpd. @Freek: if you don't usually use IPv6 (or if you even can't), can you add a "-4" option to your "dig smtp.gmail.com" command? It helped for my case. HTH, cheers. l8er manfred
participants (5)
-
Dave Howorth -
David T-G -
Freek de Kruijf -
kschneider bout-tyme.net -
Manfred Hollstein