httpdmng taking all sites down for a few minutes

[brainforge]

/usr/local/psa/admin/bin/httpdmng --reconfigure-domain mydomain.com

(a) This command took all sites down while it was running - not just mydomain.com
(b) It took a few minutes - I seem to recall the old /usr/local/psa/admin/bin/websrvmng was a lot faster (but I had not used it for a few weeks before upgrading to Plesk 10).

The same thing happens when adding a customer through the Plesk panel.
Again I cannot recall this happening with Plesk 9.
Looking at the processes it is running things like:

/usr/bin/sw-engine -c /usr/local/psa/admin/conf/php.ini /usr/local/psa/admin/sbin/httpdmng --reconfigure-batch /usr/local/psa/tmp/httpdmng-batch-g1qYgq

 

[littlefrog]

I had created a tech ticket with Parallels
They were able to see that it is a problem more with servers that have quite a few domains
and expecially if they have custom vhosts.conf files.

I only have 300 domains and if i add an alias or domain to the server i can expect HELL for the next 20 min.

They said they are now aware of the issue and it will be fixed in some future update.

 

[brainforge]

See also: http://forum.parallels.com/showthread.php?t=108083

 

[J.Wick]

I solved this problem.

1. Go into the the subdomains table in the psa database and delete any problematic subdomain records.

2. Then rename the subdomain file folders to subdomain_temp and then recreate the subdomain in Plesk.

3. Go back and delete the new subdomain folder that is created and rename the original one back to the proper name.

Everything should work.

I must also note my dissatisfaction having to waste a few hours every time I have to upgrade Plesk. Paying $35/month for this should be flawless. Please do more internal testing before releasing patches.