Doubt when changing the default PHP version at server level

[Eduardo]

Hello and good week !

When I change the server’s default PHP version, there is a change that I think should happen and it isn’t.

Server with Ubuntu 20.04, multiphp, Nginx as webserver, HCP 1.8.3

Steps:

1. At the beginning the server has the PHP 7.3 version configured as default

2. There is a site “domain.com”, this site has the value “default” configured as “Backend Template PHP-FPM”
   At this moment:
   a) In the folder “/etc/nginx/conf.d/domains” the file “domain.com.ssl.conf” has configured: “fastcgi_pass unix:/run/php/php7.3-fpm-domain.com. sock;”
   b) The file “domain.com.conf” is in the folder “/etc/php/7.3/fpm/pool.d/”

3. Next, the default PHP version of the server is changed from 7.3 to 8.0 in HestiaCP
   (Cogs > Configure > Web Server > System PHP Version > php-8.0 > Save)

4. After changing the default PHP version of the server in the HestiaCP panel, I can see that the site “domain.com” has been moved from php-7.3 to php-8.0 in “Cogs > Configure > Web Server”)
   At this moment:
   a) The “domain.com.conf” file is still in the “/etc/php/7.3/fpm/pool.d/” folder
   b) In the folder “/etc/nginx/conf.d/domains” the file “domain.com.ssl.conf” still has configured: “fastcgi_pass unix:/run/php/php7.3-fpm-domain.com .sock;”

My question is about the latter, if the domain is now running with PHP 8.0, then the correct should be the following?:
a) The “domain.com.conf” file in the “/etc/php/8.0/fpm/pool.d/” folder
b) In the “/etc/nginx/conf.d/domains” folder the “domain.com.ssl.conf” file should have: “fastcgi_pass unix:/run/php/php8.0-fpm-domain.com. sock;”

But, if I run “v-update-web-templates” in the CLI, I found that these 2 things above now are changed and are configured with PHP 8.0:
(ie “fastcgi_pass unix:/run/php/php8.0-fpm-domain.com.sock;” and the “domain.com.conf” file is now in “/etc/php/8.0/fpm/pool.d/”)

So if I’m thinking right, is it possible that when the default version is changed in the panel; is the script that changes the PHP version needing to run “v-update-web-templates”?

[aaronkempf]

@Eduardo

• At the beginning the server has the PHP 7.3 version configured as default
• There is a site “domain.com”, this site has the value “default” configured as “Backend Template PHP-FPM”
  At this moment:
  a) In the folder “/etc/nginx/conf.d/domains” the file “domain.com.ssl.conf” has configured: “fastcgi_pass unix:/run/php/php7.3-fpm-domain.com. sock;”
  b) The file “domain.com.conf” is in the folder “/etc/php/7.3/fpm/pool.d/”

Sorry to revive an older post. But now, I’ve got some concerns as well. I’m gonna go rebuild-web-domains and see if that fixes it.

I basically have discrepencies on ‘what versions of php do I have installed’. The web frontend shows 5.6 - 8.3 (which I just added).

But I was TRYING to remove older versions of php, and i got a nasty error about them not being installed.

I DID get bit pretty hard when the Sury repo was down. But I thought that I had recovered the server from this.

Also, right now, I’m enabling and disabling FASTCGI for a site, and I keep on getting ‘out of memory’ errors, and of course errors about FASTCGI in my error logs (when FASTCGI is turned off).

I love HestiaCP. I’m looking forward to getting to the bottom of this.

Also, I changed the timeouts for my Php back DOWN to 2 minutes. Earlier I had it setup as 20 minutes. Once upon a time, I was trying to upload some files (for a wordpress restore). and I kept getting timeouts.

Now, my client is complaining about ‘not being able to access a page’. I see SOME ‘out of memory errors’.

And, I changed everything to 120 seconds for the timeouts, and then I was going through all the Php settings, and somehow, it ended up being on a screen that said that I had a 60 second timeout.

And ONE of my ‘Out of memory’ errors complained about having a 256mb ram limit, but it was clearly set to 512mb through the hestiacp frontend. So, I’m still trying to test that right now. (leaf through all the pages and make sure my timeouts all say 120 seconds).

I hope I’m not wasting anyone’s time by reviving an old thread. I’m gonna go try to rebuild the web domain and see if that prevents the timeouts. I just manually added php8.3.

[Aartsie]

Thank you for testing. I’m looking forward to your results!

[aaronkempf]

@aaronkempf

lso, right now, I’m enabling and disabling FASTCGI for a site, and I keep on getting ‘out of memory’ errors, and of course errors about FASTCGI in my error logs (when FASTCGI is turned off).

I love HestiaCP. I’m looking forward to getting to the bottom of this.

Also, I changed the timeouts for my Php back DOWN to 2 minutes. Earlier I had it setup as 20 minutes. Once upon a time, I was trying to upload some files (for a wordpress restore). and I kept getting timeouts.

Now, my client is complaining about ‘not being able to access a page’. I see SOME ‘out of memory errors’.

And, I changed everything to 120 seconds for the timeouts, and then I was going through all the Php settings, and somehow, it ended up being on a screen that said that I had a 60 second timeout.

And ONE of my ‘Out of memory’ errors complained about having a 256mb ram limit, but it was clearly set to 512mb through the hestiacp frontend. So, I’m still trying to test that right now. (leaf through all the pages and make sure my timeouts all say 120 seconds).

I hope I’m not wasting anyone’s time by reviving an old thread. I’m gonna go try to rebuild the web domain and see if that prevents the timeouts. I just manually added php8.3.

I ended up decreasing that domain to 7.4 instead of 8.x. I’ve been getting some depecrated and warning informational messages in the Error log.

All I know is that HestiaCP is AWESOME fast, it’s 10x faster than any other web host that I play with throughout the week (when it comes to MB/Sec on tools.pingdom.com).