Missing [user].sock file when adding new user in DA

guidomallee

Verified User
Joined
Jun 7, 2013
Messages
15
After adding a new user in Direct Admin, Nginx returns a 502 Bad Gateway error accessing the new user's website. Checking the nginx error log of the new user shows me the following error:

Code:
2013/06/10 17:26:46 [crit] 5822#0: *1 connect() to unix:/usr/local/php55/sockets/new_user.sock failed (2: No such file or directory) while connecting to upstream, client: 177.97.202.239, server: newdomain.com, request: "GET / HTTP/1.1", upstream: "fastcgi://unix:/usr/local/php55/sockets/new_user.sock:", host: "www.newdomain.com"
2013/06/10 17:27:12 [crit] 5822#0: *6 connect() to unix:/usr/local/php55/sockets/new_user.sock failed (2: No such file or directory) while connecting to upstream, client: 177.97.202.239, server: newdomain.com, request: "GET / HTTP/1.1", upstream: "fastcgi://unix:/usr/local/php55/sockets/new_user.sock:", host: "www.newdomain.com"
2013/06/10 17:27:23 [crit] 5843#0: *2 connect() to unix:/usr/local/php55/sockets/new_user.sock failed (2: No such file or directory) while connecting to upstream, client: 177.97.202.239, server: newdomain.com, request: "GET / HTTP/1.1", upstream: "fastcgi://unix:/usr/local/php55/sockets/new_user.sock:", host: "www.newdomain.com"

Checking the folder /usr/local/php55/sockets/ confirms that new_user.sock doesn't exist. I tried creating a new user and again, the .sock file is not created. The only sock files present are admin.sock and webapps.sock.

Any workaround for this issue?
 
Ok, a simple php-fpm restart solved this issue. Hooking furthur into this I have two more questions:

1. Should it really be necesary to (manually) restart php-fpm, each time a new user is created?
2. After changing a users nginx.conf is it really necessary to restart the nginx service? This causes that all sites on my server are temporarily unavailable, right?

Thanks!
 
1. Should it really be necesary to (manually) restart php-fpm, each time a new user is created?
2. After changing a users nginx.conf is it really necessary to restart the nginx service? This causes that all sites on my server are temporarily unavailable, right?

Thanks!

(1) This is a bug, and John is aware of this.
(2) I think you can reload instead of restart -- it's perhaps a bit more friendly. However, even with the restart it's only unavailable for 1-2 seconds.
 
I have the same problem and when I want to restart php-fpm
it gives me : php-fpm unrecognized service
what should I do?
thanks
 
Back
Top