Resolved Error 522: cloudflare

[omexlu]

Hello,

From the heaven I have a problem with my websites running with cloud flare, all websites shows:
Error 522

But I don't know why, server is running and I don't have make any changes on it.

I am not on country I don't know why it's not showing that error.

A reboot has not helped. What can I do?

Please help me

 

[UFHH01]

Hi omexlu,

most likely, error "522" are temporary issues and should be "self-repaired" in a couple of minutes/hours ( depending to the worldwide sync of the depending nameservers ), especially when you didn't change anything on your server.

Pls. consider to check as well your iptables ( firewall ) and Fail2Ban ( if you use it ), because it might be, that either cloudflare IPs are banned ( did you forget to whitelist them at Fail2Ban? ), or even "localhost" or your own server IP(s) - which should as bell being whitelisted.

 

[omexlu]

Hi omexlu,

most likely, error "522" are temporary issues and should be "self-repaired" in a couple of minutes/hours ( depending to the worldwide sync of the depending nameservers ), especially when you didn't change anything on your server.

Pls. consider to check as well your iptables ( firewall ) and Fail2Ban ( if you use it ), because it might be, that either cloudflare IPs are banned ( did you forget to whitelist them at Fail2Ban? ), or even "localhost" or your own server IP(s) - which should as bell being whitelisted.

Hello,

I have checked my fail2ban, there are no ips banned. The servers and localhost has been in there on the whitelist since begin.

On Server all is running, so what can I do else?
Even on hetzner I have deactivated firewall but these doesn't help.

Thanks

Edit 2: I have paused for that domain cloud flare, and I loads but sometimes I get nginx error 500, and nothing loads very well, there are something strange on the server coming from nothing.

What can I do?

Edit: I get these message directly from cloud flare that there might be a problem on my end, but why it begins at 18 nothing has been changed here:

Looking at konz-lu.tk's 522 activity over the last 7 days I am seeing that this only started occurring around 2017-04-16T18:00:00:000Z (32 incidents so far)

Here are a pair of requests directly to your backend where you can see that there is clearly an issue

MacBook-Pro:~ mgalanter$ curl -svo /dev/null https://konz-lu.tk/ --resolve konz-lu.tk:443:46.4.84.209
* Added konz-lu.tk:443:46.4.84.209 to DNS cache
* Hostname konz-lu.tk was found in DNS cache
* Trying 46.4.84.209...
* TCP_NODELAY set
* Connection failed
* connect to 46.4.84.209 port 443 failed: Operation timed out
* Failed to connect to konz-lu.tk port 443: Operation timed out
* Closing connection 0

MacBook-Pro:~ mgalanter$ curl -svo /dev/null https://konz-lu.tk/ --resolve konz-lu.tk:443:46.4.84.209 -k
* Added konz-lu.tk:443:46.4.84.209 to DNS cache
* Hostname konz-lu.tk was found in DNS cache
* Trying 46.4.84.209...
* TCP_NODELAY set
* Connection failed
* connect to 46.4.84.209 port 443 failed: Connection refused
* Failed to connect to konz-lu.tk port 443: Connection refused
* Closing connection 0

I am seeing these errors reflected in our proxy server logs

"server": "71m126",
"flserver": "71m126",
"cacheserver": "71c126",
"severity": "error",
"pid": "3817",
"connectionId": "376449090",
"message": "upstream timed out (110: Connection timed out) while connecting to upstream",
"client": "77.47.78.205",
"request": "GET / HTTP/1.1",

"server": "71m126",
"flserver": "71m126",
"cacheserver": "71c126",
"severity": "error",
"pid": "3817",
"connectionId": "376449090",
"message": "connect() failed (110: Connection timed out) while connecting to upstream",
"client": "77.47.78.205",
"request": "GET / HTTP/1.1",

Given the recency of these errors, even though you mention that nothing has changed, it does look like something did change. I would recommend double-checking with any local technical resources you have such as your hosting provider support to see what the issue might be.

Hope that helps,
Maksim

 

[UFHH01]

Hi omexlu,

did you consider to check that apache2, nginx and your ( possible ) corresponding domain - specific PHP-FPM - service is up and running without issues/errors/problems?

"message": "upstream timed out (110: Connection timed out) while connecting to upstream",

=> points mostly to a non-working service, or misconfigured webserver - configuration files.

Given the recency of these errors, even though you mention that nothing has changed, it does look like something did change.

Correct answer.

I would recommend double-checking with any local technical resources you have such as your hosting provider support to see what the issue might be.

If the "Plesk Repair Utility" with the possible command options and the investigations of your log - files doesn't point you to your issues/errors/problems, I recommend as well the very same.

Pls. consider as well to use a "managed server" instead of a "dedicated server" with your Hetzner hosting plan, because people willing to help you to investigate your issues/errors/problems together with you have to have some entries from your log - files, informations about your configuration files and last but not least, you should consider to provide standart informations as for example:

Your current operating system
Your current Plesk version ( incl. #MU )

Plesk for Linux services logs and configuration files ( KB - article: 213403509 )​

 

[omexlu]

I don't know why but this morning all works again from alone, very strange and the first time this happend :/

 

[omexlu]

Hi omexlu,

did you consider to check that apache2, nginx and your ( possible ) corresponding domain - specific PHP-FPM - service is up and running without issues/errors/problems?

"message": "upstream timed out (110: Connection timed out) while connecting to upstream",

=> points mostly to a non-working service, or misconfigured webserver - configuration files.


Correct answer.


If the "Plesk Repair Utility" with the possible command options and the investigations of your log - files doesn't point you to your issues/errors/problems, I recommend as well the very same.

Pls. consider as well to use a "managed server" instead of a "dedicated server" with your Hetzner hosting plan, because people willing to help you to investigate your issues/errors/problems together with you have to have some entries from your log - files, informations about your configuration files and last but not least, you should consider to provide standart informations as for example:

Your current operating system
Your current Plesk version ( incl. #MU )

Plesk for Linux services logs and configuration files ( KB - article: 213403509 )​

Again the same problem :/ what can i do and where to show?

in nginx.log i found this:
2017/04/18 21:02:26 [alert] 30036#0: *799119 open socket #267 left in connection 5

and in apache this:
[Tue Apr 18 21:02:54.472523 2017] [mpm_event:error] [pid 22098:tid 139686042204032] AH00484: server reached MaxRequestWorkers setting, consider raising the MaxRequestWorkers setting

Maybe this is a problem? Maybe a DDoS because i remarked that +6000 Members was online on my community.

 

[UFHH01]

Hi omexlu,

2017/04/18 21:02:26 [alert] 30036#0: *799119 open socket #267 left in connection 5

Which nginx version do you use? Pls. post the main "nginx.conf" for further investigations.

[Tue Apr 18 21:02:54.472523 2017] [mpm_event:error] [pid 22098:tid 139686042204032] AH00484: server reached MaxRequestWorkers setting, consider raising the MaxRequestWorkers setting

Which apache2 - version do you use? Pls. post the main apache2.conf/httpd.conf for further investigations.

Could you explain why you use "mpm_event" instead of "mpm_prefork"? I remember, that you use wordpress as well on your server and with a combination apache+nginx+wordpress, the recommended usage would be PHP-FPM as PHP-handler. Pls. inform us as well about the correct domain - specific PHP-handler in use.


In addition, it is ALWAYS helpful to provide at least the following informations, when you open a thread with issues/errors/problems here in the forums:

YOUR operating system
YOUR current used Plesk version

A very good start is the output of the command ( logged in as user "root" over SSH ): plesk version​

As you might have noticed by now, your issue/error/problems appears quite regulary, which offers the possibility that logrotate configuration files are involved. Pls. consider to post as well depending apache, nginx, php - related logrotate - configuration files from "/etc/logrotate.d".

 

[omexlu]

Which nginx version do you use? Pls. post the main "nginx.conf" for further investigations.

Which apache2 - version do you use? Pls. post the main apache2.conf/httpd.conf for further investigations.

All is standart how it is providen by plesk onyx in the actual yodates. I am not in country to provide much infos at this point.

Could you explain why you use "mpm_event" instead of "mpm_prefork"? I remember, that you use wordpress as well on your server and with a combination apache+nginx+wordpress, the recommended usage would be PHP-FPM as PHP-handler. Pls. inform us as well about the correct domain - specific PHP-handler in use.

Also here is the default onyx setup nothing changed, and PHP-FPM by apache is set here.

YOUR operating system
YOUR current used Plesk version

Actual under Debian 8

As you might have noticed by now, your issue/error/problems appears quite regulary, which offers the possibility that logrotate configuration files are involved. Pls. consider to post as well depending apache, nginx, php - related logrotate - configuration files from "/etc/logrotate.d".

I will provide soon as i am at home.

But i think i kind of resource lack because i have seen +6000 Members online on the forum (maybe HTTP DDoS) who results in these error.

 

[UFHH01]

All is standart how it is providen by plesk onyx in the actual yodates. I am not in country to provide much infos at this point.

Well... and how do you expect people willing to help you to investigate the issue? By guessing? Sorry... this will not work, is too time - consuming and mostly will not lead to decent answers to solve your issue.

 

[omexlu]

Well... and how do you expect people willing to help you to investigate the issue? By guessing? Sorry... this will not work, is too time - consuming and mostly will not lead to decent answers to solve your issue.

As im said , i am on holidays and cant provide much informations at this point from here:

Copied from panel:
nginx 1.11.10-debian8.0.17040416
php 5.6.30+dfsg-0+deb8u1
php-fpm 5.6.30+dfsg-0+deb8u1
php56 5.6.30-debian8.0.17031314
php70 7.0.18-debian8.0.17041318
php71 7.1.4-debian8.0.17041413
psa-logrotate 3.8.2-debian8.0.build1700161025.15
Version Plesk Onyx v17.0.17_build1700161028.14 os_Debian 8.0
OS Debian 8.7

 

[CricStands]

Can any one help me to install CloudFlare on this domain kindly
CricStands - Cricket Is Not Just A Sport, Its a Passion

As im said , i am on holidays and cant provide much informations at this point from here:

Copied from panel:
nginx 1.11.10-debian8.0.17040416
php 5.6.30+dfsg-0+deb8u1
php-fpm 5.6.30+dfsg-0+deb8u1
php56 5.6.30-debian8.0.17031314
php70 7.0.18-debian8.0.17041318
php71 7.1.4-debian8.0.17041413
psa-logrotate 3.8.2-debian8.0.build1700161025.15
Version Plesk Onyx v17.0.17_build1700161028.14 os_Debian 8.0
OS Debian 8.7

 

[trialotto]

@CricStands

Can you be so kind to open a new topic thread?

Or, if desired, you can also send me a private message.

Kind regards.....