cant kill processes, too many lshttpd start , screws everything up

felosi

Well-Known Member
#1
This has been a big for a while now where lshttpd will restart or something and the old processes are still going causing everything to go haywire. You cannot kill the processes with killall. and you have to kill -9 the root lshttpd process over and over to get it to kill.

And since the update no other httpd server can be ran on same server. I got a media site I manage that runs apache on frontend, one ip, and its own config.

Litespeed is supposed to have its own config, a subdomain, an own ip. It is not supposed to touch apache config or intefere with apache in anyway. But that is not the case, its restarting or something and taking over apaches ip and port, something that is not even in its config.

Cant really say for sure if this is buig or what but the not killing of processes surely is
 

Lauren

LiteSpeed Staff
Staff member
#2
always use "lsws/bin/lswsctrl stop" or "service lsws stop" to stop lshttpd.
If LSWS does not use Apache httpd.conf by configured natively, and it is configured to listen on specific IP, it will not take over apache's IP : port.
 

felosi

Well-Known Member
#3
not sure, may have been something with that setup

But on the process thing, if you use the lswsctrl stop, will still take some time to stop. Should be able to kill the pids

Sometimes the lsws restarts and the old processes are still remaining and it just freezes up everything, makes sites lag badly, 503 errors, etc. And I have to go in manually and stop it all, kill processes and start again

next time I will copy the ps and show you
 

felosi

Well-Known Member
#4
Oh yeah its confirmed, new version will take over any listening ip whether its in config or not. I rolled back and fixed the problem
 

felosi

Well-Known Member
#6
This is a real bug no matter how you look at it. It does this on da servers most but does it on apache as well.

So here is how it usually goes like today. Found out all sites were down on the server, a directadmin server. I go in and , well here is copy and paste for you
[root@am1 ~]# ltr <--- (/opt/lsws/bin/lswsctrl restart)
[OK] Send SIGUSR1 to 4126
[OK] Lease License key #1 will expire in 33 days!
httpd (pid 4128?) not running
[OK] lshttpd: pid=25226.
[root@am1 ~]# ps -u apache
PID TTY TIME CMD
813 ? 00:00:00 lsphp
4129 ? 00:08:05 lshttpd
4130 ? 00:04:17 lshttpd
4842 ? 00:01:27 memcached
25245 ? 00:00:00 lshttpd
25246 ? 00:00:00 lshttpd
[root@am1 ~]# killall lshttpd
[root@am1 ~]# killall lshttpd
[root@am1 ~]# killall lshttpd
[root@am1 ~]# killall lshttpd
[root@am1 ~]# killall lshttpd
[root@am1 ~]# killall lshttpd
[root@am1 ~]# killall lshttpd
[root@am1 ~]# killall lshttpd
[root@am1 ~]# killall lshttpd
[root@am1 ~]# killall lshttpd
[root@am1 ~]# /opt/lsws/bin/lswsctrl stop
[OK] lshttpd: stopped.
[root@am1 ~]# /opt/lsws/bin/lswsctrl stop
[ERROR] lshttpd is not running.
[root@am1 ~]# ltr
[OK] Lease License key #1 will expire in 33 days!
[OK] lshttpd: pid=25353.
[OK] Lease License key #1 will expire in 33 days!
httpd (pid 25244?) not running
[OK] lshttpd: pid=25382.
[root@am1 ~]# ps -u apache
PID TTY TIME CMD
4130 ? 00:04:17 lshttpd
4842 ? 00:01:27 memcached
25245 ? 00:00:00 lshttpd
25296 ? 00:00:00 lshttpd
[root@am1 ~]# ltr
[OK] Lease License key #1 will expire in 33 days!
httpd (pid 25244?) not running
[OK] lshttpd: pid=25452.
[root@am1 ~]# ps -u apache
PID TTY TIME CMD
4842 ? 00:01:27 memcached
25296 ? 00:00:00 lshttpd
25457 ? 00:00:00 lshttpd
25458 ? 00:00:00 lshttpd
[root@am1 ~]# /opt/lsws/bin/lswsctrl stop
[OK] lshttpd: stopped.
[root@am1 ~]# /opt/lsws/bin/lswsctrl stop
[ERROR] lshttpd is not running.
[root@am1 ~]# ps -u apache
PID TTY TIME CMD
4842 ? 00:01:27 memcached
25457 ? 00:00:00 lshttpd
[root@am1 ~]# kill -9 25457
[root@am1 ~]# ps -u apache
PID TTY TIME CMD
4842 ? 00:01:27 memcached
25606 ? 00:00:00 lshttpd
25608 ? 00:00:00 lshttpd
[root@am1 ~]# ltr
[OK] Send SIGUSR1 to 25602
[root@am1 ~]# nano -w /usr/local/dir*/data/admin/services.status
[root@am1 ~]# ps -u apache
PID TTY TIME CMD
4842 ? 00:01:27 memcached
25606 ? 00:00:00 lshttpd
25691 ? 00:00:00 lshttpd
25692 ? 00:00:00 lshttpd
[root@am1 ~]# ltr
[OK] Send SIGUSR1 to 25689
[root@am1 ~]#
The processes cooperated better then normal but why is litespeed restarting and leaving zomebie processes? When it does this every site on the box goes down
 

felosi

Well-Known Member
#8
Sorry didnt see your reply on this But anyway, thi seems to happen more so on directadmin server. When it happens it makes server completely unresponsive. Sending login info now
 

felosi

Well-Known Member
#9
Well here is an instance of it happening on cpanel server. the sites started lagging badly and load was super high. I logged in server, here is the commands and output. This happens everyonce in a while on cpanel servers and frequently with directadmin servers. Please look into this, surely Im not the only one with this problem

[root@cp2 linux]# ps -u nobody
PID TTY TIME CMD
601 ? 00:00:00 lshttpd
602 ? 00:00:00 lshttpd
614 ? 00:00:00 admin_php
31141 ? 00:00:06 lshttpd
31142 ? 00:00:02 lshttpd
32713 ? 00:00:00 admin_php
[root@cp2 linux]# killall ls
[root@cp2 linux]# ps -u nobody
PID TTY TIME CMD
601 ? 00:00:00 lshttpd
602 ? 00:00:00 lshttpd
614 ? 00:00:00 admin_php
31141 ? 00:00:06 lshttpd
32713 ? 00:00:00 admin_php
[root@cp2 linux]# killall lshttpd
[root@cp2 linux]# killall lshttpd
[root@cp2 linux]# killall lshttpd
[root@cp2 linux]# killall lshttpd
[root@cp2 linux]# killall lshttpd
[root@cp2 linux]# killall lshttpd
[root@cp2 linux]# killall lshttpd
[root@cp2 linux]# /opt/lsws/bin/lswsctrl stop
[OK] lshttpd: stopped.
[root@cp2 linux]# killall lshttpd
[root@cp2 linux]# killall lshttpd
[root@cp2 linux]# ps -u nobody
PID TTY TIME CMD
601 ? 00:00:00 lshttpd
602 ? 00:00:00 lshttpd
614 ? 00:00:00 admin_php
[root@cp2 linux]# /opt/lsws/bin/lswsctrl stop
[ERROR] lshttpd is not running.
[root@cp2 linux]# ps -u nobody
PID TTY TIME CMD
601 ? 00:00:00 lshttpd
614 ? 00:00:00 admin_php
[root@cp2 linux]# kill -9 601
-bash: kill: (601) - No such process

Then I was finally able to restart it
 
Top