Submitted by Vince42 on Wed, 01/20/2016 - 16:12 Pro Licensee
I have an open issue with the proftpd (see: https://virtualmin.com/node/38759).
First of all I have a problem understanding, which status is displayed in the System Information as described in the forum topic, second I am currently facing a strange problem that the FTP daemon is quitting every few days and I cannot figure out why. The latter topic might be off topic, but I would really highly appreciate any help solving this mystery as well, as I am sick of restarting the daemon manually all the time ... :)
PS: This has not happened on my old server with Virtualmin / Webmin, but I copied the setup directly, so there should be no difference.
Submitted by JamieCameron on Wed, 01/20/2016 - 21:16 Comment #1
You should check the log file
/var/log/messagesfor any reference to or messages from
proftpdto see why it is quitting.
Submitted by andreychek on Thu, 01/21/2016 - 09:50 Comment #2
Additionally, you may also want to take a peek at the output of the command "dmesg | tail 100", to see if you see any kernel error messages. If you're seeing any resource problems, it could display them there (such as OOM Killer, which can occur in instances of low RAM).
Submitted by Vince42 on Thu, 01/21/2016 - 16:50 Pro Licensee Comment #3
just fyi: I checked /var/log/proftpd/proftpd.log and found this
2016-01-18 06:28:56,691 x3.eec.de proftpd x3.eec.de: ProFTPD killed (signal 15)
2016-01-18 06:28:56,726 x3.eec.de proftpd x3.eec.de: ProFTPD 1.3.5rc3 standalone mode SHUTDOWN
Google led me to Stack Overflow and I applied that solution by adding
--retry=1in the init.d script - keep fingers crossed. :)
Thank you both very much!
Submitted by JamieCameron on Thu, 01/21/2016 - 21:07 Comment #4
That message suggests that proftpd was intentionally shut down.
Submitted by Vince42 on Fri, 01/22/2016 - 14:49 Pro Licensee Comment #5
it is true, proftpd has been intentionally shut down everytime: logrotate restarts the service, but proftpd is somehow "slow" and stays dead - therefore they say that adding the retry option should fix the issue. I hope that this information will help you as well.
Submitted by JamieCameron on Sat, 01/23/2016 - 01:47 Comment #6
Odd that it doesn't re-start after log rotation. Is your system perhaps low on RAM?
Submitted by Vince42 on Sat, 01/23/2016 - 06:44 Pro Licensee Comment #7
It seems that there is a known issue with restarting proftd on some systems ... it is certainly not due to low resources as my server is ten times the resources than my old server aka "super powered" ... maybe just some strange "glitch", but I never experienced such behaviour from Linux ...
Just to give you an impression:
Ubuntu 14.04 - ProFTPd stops unexpectedly, easy fix ...
edit the service file, to add a retry and sleep ...
Find this line...
start-stop-daemon --stop --signal $SIGNAL --quiet --pidfile "$PIDFILE"
and replace with ...
start-stop-daemon --stop --signal $SIGNAL --retry 1 --quiet --pidfile "$PIDFILE"
Restart the service and the problem should go away.
Submitted by Vince42 on Sat, 01/23/2016 - 15:24 Pro Licensee Comment #9
... exactly what I researched and posted under #3 - thank you for the confirmation!
Submitted by hostile on Sat, 02/13/2016 - 01:56 Comment #10
didnt worked out on debian 8, still gets killed
Submitted by andreychek on Sat, 02/13/2016 - 10:11 Comment #11
Yeah we've heard a few reports of this occurring on a few different distros. Though it doesn't occur on all of them which makes it tricky to troubleshoot.
It does sound like there's a packaging bug in the case of some of these, which we're hoping will be resolved soon.
In the meantime, you may want to setup monitoring using Webmin -> Others -> System and Server Status.
Using that, you can ensure that ProFTPd continues running even if it's killed off during the package rotation.
Submitted by AustinTX on Mon, 04/18/2016 - 16:16 Comment #12
I've had this problem for a year on my server (which is now Webmin 1.791 on Ubuntu 14.04.1). I too see in my log:
/var/log/proftpd/proftpd.log.1:2016-04-17 01:39:36,185 ns1 proftpd ns1: ProFTPD killed (signal 15)
/var/log/proftpd/proftpd.log.1:2016-04-17 01:39:37,898 ns1 proftpd ns1: ProFTPD 1.3.5rc3 standalone mode SHUTDOWN
/var/log/proftpd/proftpd.log.1:2016-04-17 01:41:13,820 ns1 proftpd ns1: ProFTPD 1.3.5rc3 (devel) (built Fri Dec 20 2013 18:05:41 UTC) standalone mode STARTUP
/var/log/proftpd/proftpd.log.1:2016-04-17 07:06:34,769 ns1 proftpd ns1: ProFTPD killed (signal 15)
/var/log/proftpd/proftpd.log.1:2016-04-17 07:06:35,318 ns1 proftpd ns1: ProFTPD 1.3.5rc3 standalone mode SHUTDOWN
The STARTUP is me manually restarting from Virtualmin's "System Information | Status" icon. Btw- if I start proftpd from the "Servers" module, the "System Information" page will still show it unstarted. Bug?
I'll give the "--retry 1" solution a try. If it doesn't work, I'll try the FOREVER command. Btw- would there be any harm in increasing retry to a higher number? Or does "1" just mean "yes" here?
Submitted by AustinTX on Mon, 05/09/2016 - 15:12 Comment #13
Ok, three Sundays have gone by, and ProFTPd no longer dies. All I did was add the "--retry 1" statement to /etc/init.d/proftpd as above.
I think we have a confirmed solution for my platform.
Ubuntu Linux 14.04.1, Webmin 1.795, Virtualmin 5.02.gpl