Websites destroyed after nightly Update

[Squeeze999]

After the Panel update last night none of our customers websites looks like before. No images, no stylesheet is working.
I noticed this on three Plesk Panel Servers

Please help! It is urgent

 

[IgorG]

Any details, results of troubleshooting?
Sorry, but "It doesn't work! Please help urgently!" is not enough for problem investigation.

 

[Squeeze999]

Hi Igor,

sorry for my Panic...
On all the PHP driven sites images and css styles are missing. Restarting the Server did not help. Panel access is possible.
But I meanwhile found a workaround: I have to switch php execution from Apache module to fastCGI and back. After that, the site is displayed again.
I will have to do this on around 50 sites.

Was there any change in PHP with the last update?

 

[IgorG]

I see no any changes in PHP with the latest MU#14 - http://download1.parallels.com/Plesk/PP12/parallels-plesk-12.0-for-linux-change-log.html#1218-u14

 

[Squeeze999]

Hi Igor,

it seems, as if it is no php issue. Pure HTML sites are affected, too. Pure HTML sites deliver a 404.
There has to be a change in hosting settings to reactivate the correct display of the sites.

You see the behaviour on www.trauerredner-remstal.de
It is a TYPO3-site, that is still in the state after the update last night.

Tell me, when you die see it, I'd like to apply my Workaround

 

[IgorG]

Yes I see the issue. Actually, logs of webserver should be investigated - why part of content doesn't handled.

 

[RedEnzian]

a

/usr/local/psa/admin/bin/httpdmng --reconfigure-all has solved the problem for me.

additionally ssl certs did not work before. You can also fix the problem host by host when changing a value in the hostingsettings, save it and than switch bag

 

[Squeeze999]

@IgorG: Logs look clean, except the certificate warning. Do you need any? Please let me know

@RedEnzian: thanks for the abbreviation . I already found the long way workaround with the hosting settings, but with > 50 domains on a Server, you likely loose the overview...

 

[MarceauC]

I had the same problem this night!

It seems that this was due to a nginx misconfiguration.
I solved the problem editing nginx settings for each domain under the "hosting settings" tab.

But I would understand why this happened with the Plesk upgrade 12.0.18#14...

Edit : I've 2 main plesk servers that have been upgraded this night. The server that was crashed has PHP 5.4.4 from atomic repo, while the other did not (and has not crashed).
Should I disable it from /etc/yum.repos.d/atomic.repo to prevent upgrade problem?

 

[Arusnet]

Same

Hi, I had the same problem. I've solved what MarceauC commented. I have php 5.3.3 and Plesk 12.0.18 # 14.
I had to change a site by site to "default certificate" in "hosting config" I just hope it does not affect the security of the site.

regards,

 

[ColetteC]

We have got the same problem too. All our styles and css is no longer showing. Please advise of what we need to do.

 

[Squeeze999]

There are two possible workarounds:

a: Change any setting in the Hosting Settings, save it and turn it back. I did it with a switch from P>HP as Apache Module to FastCGI and back.
b: As RedEnzian wrote above: run /usr/local/psa/admin/bin/httpdmng --reconfigure-all on your machine. It should work, too

Addendum: Workaround b does not solve the problem. You have to to do Workaround a for every Domain and subdomain with a website.

@Parallels: please try to find a solution, as it is not really amusing, to manually correct that issue. And, as you may have noticed, I am not the only customer suffering from that Problem.

 

[Atramasis]

Same problem here!

For me /usr/local/psa/admin/bin/httpdmng --reconfigure-all worked. Thanks.

 

[IgorG]

Problem was reported as RT #1957376 and under developers investigation now. Thank you for information.

 

[Gracjan]

Same problem after last update. All my websites are crashed! ;(
/usr/local/psa/admin/bin/httpdmng --reconfigure-all This helping me.

 

[IgorG]

Bug created (PPPM-1930 for your reference) We will publish new microupdate immediately when fix will be ready.

 

[Gracjan1]

I'm guessing or there is the same problem after today's update?
On all of my sites:
502 Bad Gateway
nginx

 

[UFHH01]

A report of "502 Bad Gateway" is not really a good way of investigating eoors/issues/problems, because the cause can vary from PHP issues, over nginx failures, to php5-fpm problems, to firewall problems.... please be so kind to provide some logs when reporting, so that people willing to help may suggest possible work-arounds or solutions.

Parallels Plesk Panel for Linux services logs and configuration files ( KB - article: 111 283 )​

Yes, there was an update on the 13.11. - but on servers which I administrate ( Ubuntu /Debian ), there was no issue at all after the MU - update.