Resolved Got a 502 error after disabling several apache modules.

[Ivan_K]

I was disabling one ore two Apache modules. And got a 502 error.
I have an error from Nginx in the logs:

188661#0: *408 connect() failed (111: Connection refused) while connecting to upstream

My attempts to get the server back up and running, even by enabling ALL modules, were unsuccessful.

How i can reset apache modules list?

PS. FPM used via Apache and Nginx used as proxy.

 

[Ivan_K]

ADD: my attempt to use FPM served by nginx had the same result (502 error).

 

[WebHostingAce]

Make sure your Apache is running at the moment.

 

[Ivan_K]

It's running

 

[Ivan_K]

I havn't the unixid apache module (which used for fcgid module) in cofig. How i can restore it?

 

[IgorG]

I havn't the unixid apache module (which used for fcgid module) in cofig. How i can restore it?

It is part of httpd package from OS vendor. For example:

# rpm -qf /etc/httpd/modules/mod_unixd.so
httpd-2.4.6-97.el7.centos.x86_64

 

[Ivan_K]

It is part of httpd package from OS vendor. For example:

# rpm -qf /etc/httpd/modules/mod_unixd.so
httpd-2.4.6-97.el7.centos.x86_64

this problem on the new server under Ubuntu 20 LTS.
apt reinstall apache2 has not resolve this problem

/usr/lib/apache2/modules# ls -a
. mod_authz_owner.so mod_env.so mod_mime.so mod_rewrite.so
.. mod_authz_user.so mod_expires.so mod_mpm_event.so mod_security2.so
httpd.exp mod_autoindex.so mod_ext_filter.so mod_mpm_prefork.so mod_sed.so
mod_access_compat.so mod_brotli.so mod_fcgid.so mod_mpm_worker.so mod_session_cookie.so
mod_aclr.so mod_bucketeer.so mod_file_cache.so mod_negotiation.so mod_session_crypto.so
mod_actions.so mod_buffer.so mod_filter.so mod_proxy_ajp.so mod_session_dbd.so
mod_alias.so mod_cache_disk.so mod_headers.so mod_proxy_balancer.so mod_session.so
mod_allowmethods.so mod_cache.so mod_heartbeat.so mod_proxy_connect.so mod_setenvif.so
mod_asis.so mod_cache_socache.so mod_heartmonitor.so mod_proxy_express.so mod_slotmem_plain.so
mod_auth_basic.so mod_case_filter_in.so mod_http2.so mod_proxy_fcgi.so mod_slotmem_shm.so
mod_auth_digest.so mod_case_filter.so mod_ident.so mod_proxy_fdpass.so mod_socache_dbm.so
mod_auth_form.so mod_cern_meta.so mod_imagemap.so mod_proxy_ftp.so mod_socache_memcache.so
mod_authn_anon.so mod_cgid.so mod_include.so mod_proxy_hcheck.so mod_socache_redis.so
mod_authn_core.so mod_cgi.so mod_info.so mod_proxy_html.so mod_socache_shmcb.so
mod_authn_dbd.so mod_charset_lite.so mod_lbmethod_bybusyness.so mod_proxy_http2.so mod_speling.so
mod_authn_dbm.so mod_data.so mod_lbmethod_byrequests.so mod_proxy_http.so mod_ssl.so
mod_authn_file.so mod_dav_fs.so mod_lbmethod_bytraffic.so mod_proxy_scgi.so mod_status.so
mod_authn_socache.so mod_dav_lock.so mod_lbmethod_heartbeat.so mod_proxy.so mod_substitute.so
mod_authnz_fcgi.so mod_dav.so mod_ldap.so mod_proxy_uwsgi.so mod_suexec.so
mod_authnz_ldap.so mod_dbd.so mod_log_debug.so mod_proxy_wstunnel.so mod_sysenv.so
mod_authz_core.so mod_deflate.so mod_log_forensic.so mod_ratelimit.so mod_unique_id.so
mod_authz_dbd.so mod_dialup.so mod_lua.so mod_reflector.so mod_userdir.so
mod_authz_dbm.so mod_dir.so mod_macro.so mod_remoteip.so mod_usertrack.so
mod_authz_groupfile.so mod_dumpio.so mod_md.so mod_reqtimeout.so mod_vhost_alias.so
mod_authz_host.so mod_echo.so mod_mime_magic.so mod_request.so mod_xml2enc.so

 

[Ivan_K]

Screen from panel added

 

[Ivan_K]

Igor, i need help as soon as possible. Help plz!

 

[weltonw]

Open a support ticket with Plesk if you need urgent support.

In the future, you obviously shouldn't randomly disable Apache modules that you aren't sure of. That's basically begging for trouble.

First, I'd see if apache is working. Can you connect directly to the Apache port? (Usually 7081).

What if you turn off Apache for a site? Does that fix the error?

Are there any errors in Apache error logs?

 

[WebHostingAce]

In addition to to the above replies. You can also try,

- Restart both nginx and apache from the command-line and see if they are restarting without any issues.
- then try #plesk repair web

The error "188661#0: *408 connect() failed (111: Connection refused) while connecting to upstream" means Apache is not responding to NGINX requests

 

[WebHostingAce]

Sorry, I just read this thread properly. It seems unixd in not a default Apache Module for Plesk/Ubuntu.

How to enable/disable Apache modules shipped with Plesk

Applicable to: Plesk Onyx for Linux Question How to enable/disable Apache modules shipped with Plesk?How to check what Apache modules are enabled/available? Answer Note: This guide is intended f...

support.plesk.com

 

[Ivan_K]

The error "188661#0: *408 connect() failed (111: Connection refused) while connecting to upstream" means Apache is not responding to NGINX requests

yes, this error are in the nginx logs, but why?

 

[Ivan_K]

Problem solved

All websites hosted in Plesk are not accessible over HTTPS: 502 Bad Gateway or ERR_CONNECTION_REFUSED

Applicable to: Plesk for Linux Symptoms Websites hosted in Plesk are loading over HTTP, but do not work over HTTPS: CONFIG_TEXT: 502 Bad Gateway nginx CONFIG_TEXT: ERR_CONNECTION_REFUSED The f...

support.plesk.com