After Ubuntu 14.04 -> 16.04 upgrade, pypMyAdmin retuns 502 error on nginx

Question

So I just upgraded my VPS from 14.04 to 16.04, the upgrade went well for most parts, but MySQL failed to update. I needed to manually remove mariadb-client first, and then I was able to install MySQL again, and everything on my website is now working fine, except for phpMyAdmin. Whenever I open the site.com/phpmyadmin page it displays an nginx 502 error, bad gateway.

The relevant parts of my nginx config file looks like this:

location ~ \.php$
{
    fastcgi_index index.php;
    fastcgi_keep_conn on;
    include /etc/nginx/fastcgi_params;
    fastcgi_pass unix:/var/run/php/php7.1-fpm.sock;
    fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
}

location /phpmyadmin
{
    root /usr/share;

    location ~ \.php$
    {
        include php.conf;
    }

    location ~* ^/phpmyadmin/(.+\.(jpg|jpeg|gif|css|png|js|ico|html|xml|txt))$
    {
        root /usr/share/;
    }
}

I looked around the net to see if anyone faced similar issues, but in most cases the solution was an incorrect php-fpm setting.

In my case it seems to be working well:

root@server:/etc/nginx/conf.d-enabled# service php7.1-fpm status
php7.1-fpm start/running, process 28542

root@server:/etc/nginx/conf.d-enabled# ls -l /var/run/php/php7.1-fpm.sock
srw-rw-rw- 1 www-data www-data 0 Jan  9 15:01 /var/run/php/php7.1-fpm.sock

PhpMyAdmin worked well before the upgrade to 16.04, and apart from removing mariadb-client nothing else happened besides the automatic process.

Any ideas where should I dig to find a solution? Thanks in advance for any help!

Sukhjinder Singh · Answer 1 · 2020-01-09T17:48:57.983

0

Change the fastcgi pass line

fastcgi_pass unix:/var/run/php/php7.0-fpm.sock

edited Jan 09 '20 at 17:48

answered Jan 09 '20 at 15:53

Sukhjinder Singh

1,994
2
9
17

My server uses php7.1 and not php5, so why would I do that? – LBandy Jan 09 '20 at 15:59
check your php-fpm version – Sukhjinder Singh Jan 09 '20 at 15:59
I did and I already included the relevant parts in my original question. php7.1-fpm is running on my server and is used to serve the rest of the website, only phpMyAdmin is having issues. – LBandy Jan 09 '20 at 16:06
Please check your php logs. – Sukhjinder Singh Jan 09 '20 at 16:27
Okay so the issue seem to be the following: 52 connect() to unix:/var/run/php5-fpm.sock failed (2: No such file or directory) while connecting to upstream Now I just need to figure out where the default configuration is coming from. – LBandy Jan 09 '20 at 17:37
try unix:/var/run/php/php7.0-fpm.sock – Sukhjinder Singh Jan 09 '20 at 17:49
I'm setting the correct fpm in my nginx config file. But for some reason when trying to access phpMyAdmin the server is looking for unix:/var/run/php5-fpm.sock instead. I don't have php5 installed anymore, so I'm wondering where this configuration is coming from. – LBandy Jan 09 '20 at 18:01
Probably `php.conf`. – womble Jan 09 '20 at 22:58

score 0 · Accepted Answer · answered Jan 09 '20 at 18:12

So the solution was that in nginx config the link "include php.conf;" parsed a php.conf file which persumably got updated during the process. It was indeed a php-fpm issue, as it contained the "fastcgi_pass unix:/var/run/php/php5-fpm.sock" instead of "fastcgi_pass unix:/var/run/php/php7.1-fpm.sock":

connect() to unix:/var/run/php5-fpm.sock failed

After changing it I received the following error:

upstream sent too big header while reading response header from upstream, client

Adding the following lines there fixed the issue, I'm now able to access phpMyAdmin:

fastcgi_buffers 16 16k; 
fastcgi_buffer_size 32k;

After Ubuntu 14.04 -> 16.04 upgrade, pypMyAdmin retuns 502 error on nginx

2 Answers2