Plesk Login: 500 Internal Server Error

[johnrt]

I have a new Plesk instance running on AWS. Everything with Plesk itself has been fine. Today I try to login and I get 500 - Internal Server Error.

Logging in at https://x.x.x.x:8443 as usual.

The websites are fine and I can SSH in OK.

None of the other threads on this seem to shine much light on the problem.

First question:
1) Where are the log files for Plesk so I can see what the 500 error is all about?

 

[johnrt]

Sometimes when I refresh the login page, I will get:
503 - Service Not Available

 

[IgorG]

http://kb.parallels.com/111283

 

[johnrt]

Error Logs

Not certain if any of these logs tell us anything:

us-east-1c-i-a16a0ede [root:sw-cp-server]$ tail error_log
2012-12-06 20:47:02: (mod_fastcgi.c.3962) pid 30372 5 not found: No child processes
2012-12-06 20:47:03: (mod_fastcgi.c.3962) pid 30372 5 not found: No child processes
2012-12-06 20:47:04: (mod_fastcgi.c.3962) pid 30372 5 not found: No child processes
2012-12-06 20:47:05: (mod_fastcgi.c.3962) pid 30372 5 not found: No child processes
2012-12-06 20:47:06: (mod_fastcgi.c.3962) pid 30372 5 not found: No child processes
2012-12-06 20:47:07: (mod_fastcgi.c.3962) pid 30372 5 not found: No child processes
2012-12-06 20:47:08: (mod_fastcgi.c.3962) pid 30372 5 not found: No child processes
2012-12-06 20:47:09: (mod_fastcgi.c.3962) pid 30372 5 not found: No child processes
2012-12-06 20:47:10: (mod_fastcgi.c.3962) pid 30372 5 not found: No child processes
2012-12-06

us-east-1c-i-a16a0ede [root:sw-cp-server]$ cd /usr/local/psa/admin/logs
us-east-1c-i-a16a0ede [root:logs]$ tail httpsd_access_log
::ffff:71.63.4.52 23.23.89.109:8443 - [06/Dec/2012:08:04:15 +0700] "GET /smb/service-instance-aps/index/resourceId/4804c37e-268f-472a-8157-7d484a5b60b4/entry/9 HTTP/1.1" 302 0 "https://23.23.89.109:8443/smb/service/index/resourceId/4804c37e-268f-472a-8157-7d484a5b60b4" "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:17.0) Gecko/20100101 Firefox/17.0"
::1 localhost:8443 - [06/Dec/2012:08:31:52 +0700] "POST /check-plesk.php HTTP/1.1" 200 794 "-" "Dr.Web Daemon/Plesk Edition"
::1 localhost:8443 - [06/Dec/2012:09:31:43 +0700] "POST /check-plesk.php HTTP/1.1" 200 770 "-" "Dr.Web Daemon/Plesk Edition"
::1 localhost:8443 - [06/Dec/2012:11:01:15 +0700] "POST /check-plesk.php HTTP/1.1" 200 784 "-" "Dr.Web Daemon/Plesk Edition"
::1 localhost:8443 - [06/Dec/2012:12:02:33 +0700] "POST /check-plesk.php HTTP/1.1" 200 790 "-" "Dr.Web Daemon/Plesk Edition"
::ffff:66.249.73.112 www.onsiteclarity.com:8443 - [06/Dec/2012:15:14:57 +0700] "GET /robots.txt HTTP/1.1" 404 345 "-" "Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)"
::ffff:66.249.73.112 www.onsiteclarity.com:8443 - [06/Dec/2012:15:14:57 +0700] "GET /javascript/common.js?plesk_version=psa-11.0.9-110120608.16 HTTP/1.1" 200 2225 "-" "Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)"
::ffff:71.0.82.131 23.23.89.109:8443 - [06/Dec/2012:19:46:10 +0700] "GET /login_up.php3 HTTP/1.1" 500 369 "-" "Mozilla/5.0 (Windows NT 6.1; rv:16.0) Gecko/20100101 Firefox/16.0"
::ffff:71.0.82.131 23.23.89.109:8443 - [06/Dec/2012:19:46:10 +0700] "GET /favicon.ico HTTP/1.1" 200 1150 "-" "Mozilla/5.0 (Windows NT 6.1; rv:16.0) Gecko/20100101 Firefox/16.0"
::ffff:71.0.82.131 23.23.89.109:8443 - [06/Dec/2012:19:46:17 +0700] "GET / HTTP/1.1" 503 369 "-" "Mozilla/5.0 (Windows NT 6.1us-east-1c-i-a16a0ede


[root:logs]$ tail panel.log

2012-12-03T08:54:40-05:00 ERR (3): File sharing list retrieval failed: sharing type=private, error=Internal Server Error
2012-12-03T08:54:44-05:00 ERR (3): File sharing list retrieval failed: sharing type=private, error=Internal Server Error
2012-12-03T08:54:55-05:00 ERR (3): File sharing list retrieval failed: sharing type=private, error=Internal Server Error
2012-12-03T08:56:17-05:00 ERR (3): File sharing list retrieval failed: sharing type=private, error=Internal Server Error
2012-12-03T08:56:24-05:00 ERR (3): File sharing list retrieval failed: sharing type=private, error=Internal Server Error
2012-12-05T20:00:38-05:00 ERR (3): Response from sappmng in stderr: TmpFile autoclean failed: No such file or directory

2012-12-05T20:03:23-05:00 ERR (3): Response from sappmng in stderr: PHP Notice: Undefined variable: oldsettings_modify_hash in /usr/local/psa/var/apspackages/apscatalogemd0Nl.zipe3685d8f-2370-26/cache/scripts/configure on line 71

us-east-1c-i-a16a0ede [root:logs]$

 

[johnrt]

PSA Restart

Did a "psa restart" and still get the errors.

 

[johnrt]

Resolved

The problem in this case was due because the file system was full. Had a log file that filled up the 200G. Deleted the log file, restarted the process that was holding onto the files (lsof +L1), and restarted Plesk (/etc/init.d/sw-cp-server restart) and that seem to resolve the issue.

To Parallels team responsible for the AWS instance of Plesk, it would be helpful that during the initial configuration of the server, you force the user to pick the time zone of the server and use that to set the timezone in php.ini. That really filled up the error log.

 

[puntable]

The problem in this case was due because the file system was full. Had a log file that filled up the 200G. Deleted the log file, restarted the process that was holding onto the files (lsof +L1), and restarted Plesk (/etc/init.d/sw-cp-server restart) and that seem to resolve the issue.

To Parallels team responsible for the AWS instance of Plesk, it would be helpful that during the initial configuration of the server, you force the user to pick the time zone of the server and use that to set the timezone in php.ini. That really filled up the error log.

Wow, you saved my life with this. I had same problem, found the issue to be a very big error log file. Deleted file, restarted server and it worked again.

 

[MislavO]

I had same problem before some time with daily crons, they never finished due to size of the logs (40-50GB) and CPU load was every day very high.

Because of that I created this simple script and i placed it in /var/www/vhost and my file name is file_size_check.sh (replace this by your needs):

#!/bin/bash

#send output to this email
email=put your email here - report with output will be send to your email

#date format - 2012-10-03
dateBackup=`date +%F`

# 2012-10-01 - day before yesterday
dateBefore=`date +%F --date='-2 day'`

find . -size +512M > file_size_$dateBackup.txt

filenameMO=file_size_$dateBefore.txt

#sending email report
mail -s "Daily file size check on YOUR_SERVER_HOSTNAME - `date +'%F'`" "$email" < file_size_$dateBackup.txt

function delete_old_file_size {
if [ -f $filenameMO ]; then
rm $filenameMO
fi
}

#calling the function
delete_old_file_size

And setup cron (if you've placed your file in /var/ww/vhost) like this:

0 6 * * * cd /var/www/vhosts && sh file_size_check.sh

This is just a simple script that will run through all vhost structure (means through all possible domains), will report for any files larger then 512MB (you can also adjust this by your needs) and send report to the email you've specified. Since I'm running this script on many servers, I've put in the script email subject MY SERVER HOSTNAME so i know exactly on what server are big files.

Please note:
- I keep this report files only 2 days, after that they are deleted from the server
- if you don't know what you're doing, please change only bold (put your email, change size by your needs, put your hostname and name of the file should be file_size_check.sh)
- I'm running this scripts on Plesks that have each 300+ domains and this scan complete in less then a 30 seconds, so the script itself won't do any harm to your server and won't cause any high load on the server
- if your report is long, ignore everything and just look for the: <yourdomain>/statistics/logs (yes, we're checking only logs, but sometimes can be helpfull for other things).

 

[Faris Raouf]

That's a very useful script for any server.

Thank you for posting it and for explaining what it does etc so that people can modify it as required.