• If you are still using CentOS 7.9, it's time to convert to Alma 8 with the free centos2alma tool by Plesk or Plesk Migrator. Please let us know your experiences or concerns in this thread:
    CentOS2Alma discussion

Default page on working domain after backup requires httpd restart

D

DeanW

Guest
I think this might have something to do with the backup process. It seems that ONE of my two domains on the server does not recover. My client reported the site went down at exactly the backup time, but it never came back up even 7 hours later. I verified the other site still worked and this site didn't. So I assume this is a problem with turning the domain back on after a backup. Note that this didn't happen with a manual backup and doesn't seem to happen every time, just once every few weeks (its a weekly script)

Plesk Version:
psa-7.5.4-fc2.build75050824.12

Backup Version:
psa-bu-7.5.4-fc2.build75050926.17

Backup Command:
/usr/local/psa/bin/psadump -f - --nostop --do-not-dump-logs | gzip | split -b 1000m - dump-`hostname`.

Restarting apache seems to fix it. I have added a "/sbin/service httpd restart" after the Plesk backup to hopefully resolve the issue, but wish it never had to happen in the first place.

If anyone has the same issue and/or any ideas please let me know.

Regards, DW

UPDATE: This is in the logs...

Jan 29 04:00:06 my.server.com psadump: Can't locate XML/Parser/Style/Dom.pm

UPDATE: I manually updated the XML::Dom perl module check out this post and the message seems to have disappeared. Not sure if I'll ever know about the reason a domain might not be turned back on. :)
 
I have created a small script file that restarts apache/mysql/psa imediately after creating the backup.

This script is run with a cron task on a daily basis.
 
I seem to have experienced an httpd crash NOT at backup time this time. I saw the message "MaxClients" reached in the logs and increased this in the config file. Of course, I don't think this will fix it either... and don't know why it would freeze completely.....

DW
 
My recent troubles were, I believe, caused by a runaway client script filling up their error_log to 18GB resulting in the /home folder being unwritable. The server would take a connection but wouldn't show anything...

Due to a short temporary error when restarting httpd manually with 'httpd restart' I added a 5 second delay between httpd stop and httpd start proceeding my backup script.

DW
----
Unlimited email lists with a PHP script on your web site!
http://listmailpro.com
 
Back
Top