cPanel Quality Control and httpd.conf corruption

[wswd]

After submitting a support ticket on this issue, I'm just trying to figure out what has been happening to cPanel quality control as of late.

First we have this thread: http://forums.cpanel.net/f5/changelog-never-updated-why-380631.html

It has been broken for well over a year. Took 5 months to fix after recognizing that it was in fact a problem.


Now, we have httpd.conf corruption in the latest update. According to the ticket I just submitted:

It looks like this issue has been reported in internal ticket #980117. It seems this may happen after a forced update upon various server activities such as adding domains, removing domains, adding accounts, removing accounts, changing a site's IP address, and adding or removing SSL certificates. Basically, any activity that causes apache configuration to be reconfigured on the fly within WHM or cPanel.

In the meantime, you should be able to resolve this by running "/scripts/rebuildhttpdconf".

So, we're expected to run this on every single server, every single time an account or subdomain is created, removed, add/remove SSL certificate, etc., across our entire network...

Unacceptable. This is entirely unprofessional, and how this ever made it out of Edge or whatever, I'll never know. Has the quality control here at cPanel really gotten so bad to where things like this (and the thread I listed above) are allowed to happen?

 

[ethical]

yea this needs to be fixed asap, as per the reply i got from support, the internal case is #98681

 

[wswd]

yea this needs to be fixed asap, as per the reply i got from support, the internal case is #98681

Your internal Case# is correct. They messed up on my initial response and later corrected it.

Just not sure how something like this can happen, with all the testing that is supposedly done before a release. Doesn't make sense.

 

[Infopro]

@wswd, that thread about the changelog servers not syncing proper has nothing to do with your cPanel.
@ethical, your ticket has nothing to do with this thread at all. Not sure why you even posted here.

 

[ethical]

@wswd, that thread about the changelog servers not syncing proper has nothing to do with your cPanel.
@ethical, your ticket has nothing to do with this thread at all. Not sure why you even posted here.

@infopro

because this is THE EXACT same issue i am having, wswd was simply referencing another issue as well, but his comment was about the httpd corruption and having to rebuild httpdconf all the time.... same issue....

Now, we have httpd.conf corruption in the latest update. According to the ticket I just submitted:
It looks like this issue has been reported in internal ticket #980117. It seems this may happen after a forced update upon various server activities such as adding domains, removing domains, adding accounts, removing accounts, changing a site's IP address, and adding or removing SSL certificates. Basically, any activity that causes apache configuration to be reconfigured on the fly within WHM or cPanel.

In the meantime, you should be able to resolve this by running "/scripts/rebuildhttpdconf".

 

[Infopro]

Greetings, please accept my apologies. I located a support ticket ID #98681 that you mentioned above early this morning, and did not notice that this ticket was not yours, but someone else's, regarding a quota issue. Hence my misplaced comment to you.

The ID you mentioned, Case #98681 is related to proxy sub domains not being updated with the released patch, properly every time.

Can you please post the ticket ID that you just opened you mention?


Thanks!

 

[cPanelKenneth]

We are currently assessing the quality of the proxy domains feature. As you point out it has some unacceptable flaws right now.

 

[ethical]

Greetings, please accept my apologies. I located a support ticket ID #98681 that you mentioned above early this morning, and did not notice that this ticket was not yours, but someone else's, regarding a quota issue. Hence my misplaced comment to you.

The ID you mentioned, Case #98681 is related to proxy sub domains not being updated with the released patch, properly every time.

Can you please post the ticket ID that you just opened you mention?


Thanks!

oh np

ticket 4872511

 

[wswd]

@wswd, that thread about the changelog servers not syncing proper has nothing to do with your cPanel.

Never said it had to do specifically with MY cPanel, rather cPanel in general. It points toward my title, which really should be changed to, "What the heck happened to quality control around here?" I was just curious as to why it would take over a year to resolve such an easy thing to fix.

Anyway, back to this thread. Not sure about the proxy domains. I personally think it has more to do with IPv6, IMO, as I noted in my ticket, as the part of the httpd.conf file that is getting corrupted includes [2001:443 >, which is trying to point something to our IPv6 assignment. It is just truncating it in some very weird way. The 2001: is indeed the first part of our IPv6 assignments.

 

[wswd]

Not to mention, tomorrow will be 10 days since the problem was recognized. How could it take 10 days to publish a fix for something that, quite honestly, is a pretty major issue...especially when you have a bunch of servers?

 

[wswd]

This is lovely. .13 was released (nobody knows what's in it...since the changelog hasn't been updated again), and this fix isn't included.

 

[ethical]

This is lovely. .13 was released (nobody knows what's in it...since the changelog hasn't been updated again), and this fix isn't included.

still nothing on this issue???

 

[Shavaun]

There was a delay in the publishing of the change log due to some maintenance on the documentation server.

The Change Log now displays correctly for me on 11.42.1.13, both at http://documentation.cpanel.net/display/ALD/11.42+Change+Log and inside WHM.

Are you still experiencing the issue? If so, are you seeing the issue inside WHM or are you looking at the link above? What is the last update that you see?

 

[wswd]

There was a delay in the publishing of the change log due to some maintenance on the documentation server.

The Change Log now displays correctly for me on 11.42.1.13, both at http://documentation.cpanel.net/display/ALD/11.42+Change+Log and inside WHM.

Are you still experiencing the issue? If so, are you seeing the issue inside WHM or are you looking at the link above? What is the last update that you see?

The changelog issue has been fixed...yes. The topic of this thread (httpd.conf corruption) has not. Quite honestly, it has gotten completely ridiculous. It has now been a week, an interim release (.13) has come out, and nothing has been done to address this problem.

The temporary "fix" you guys gave me via ticket is to run /scripts/rebuildhttpdconf as a /scripts/postwwwacct hook. That's great, but that only works 65-75% of the time. So 25+% of the time, when we create or remove or change an account, or subdomain, etc., or a reseller does any of those, Apache crashes completely, and we have to go in and edit the damn thing manually. You get this happening over a couple dozen servers and I think you can see why I'm so aggravated.

I don't know what exact release broke this (I'm guessing .12), but how the release could get past you guys, then beta and Edge and lord knows what else with all these problems just boggles the mind. It's almost as if absolutely no testing happens on your end before these releases are put out. Now we're stuck with having to fix broken httpd.conf files all day long, literally 24/7.

This is completely unacceptable, and even more unacceptable that it has now been 7 days and not a single hint of a fix for this has been made. What do you guys do all day? This should be a priority.

 

[cPanelMichael]

Hello

Thank you taking the time to provide us with your honest feedback. I wanted to assure you that internal case 99565 is still open. Our developers are actively working to resolve the issue. During the investigation, additional issues were found that need to be addressed in order to ensure the proxy subdomains feature works as expected. Thus, a resolution has not been pushed out as fast as we would all prefer.

Here are a list of the current workarounds that users have reported success with thus far:

1. Disable IPv6.
2. Disable proxy subdomains in "WHM > Tweak Settings".
3. Modify the Apache template.

Here is a quote from one of our support tickets on the third workaround mentioned:

The third step involves copying the main.default template to main.local and revising some entries within the file. For example, if you use Apache 2.4, you would run the following command:

cp /var/cpanel/templates/apache2_4/main.default /var/cpanel/templates/apache2_4/main.local

Then, revise the following two lines in the main.local file:

<VirtualHost[% FOREACH server_ip IN ips_in_use  %] [% "${server_ip}:80" %][% END %] 127.0.0.1:80>

<VirtualHost[% FOREACH server_ip IN ips_in_use  %] [% "${server_ip}:443" %][% END %] 127.0.0.1:443>

Change these to hard code the actual IP addresses rather than allowing cPanel to automatically populate that section in the main.local file, then run /scripts/rebuildhttpdconf and restart Apache to ensure that the changes show up correctly. The above hard-coding of the correct IPs should prevent the issue from re-surfacing until case 99565 has been fixed.

Thank you.

 

[wswd]

Thanks Michael,

I'll give all those a try. Is there any reason not a single one of those methods was mentioned in the support ticket?

 

[cPanelMichael]

Could you confirm your ticket number? I could only find a reference to the internal case number you were presented with.

Thank you.

 

[ethical]

i would like to note that #1 wont work for me since i dont have ipv6 even enabled?
#2 turning off proxy subdomains is not an option as this is basically creating another problem (users cant access the cpanel if they cant get there from the URL!
#3) Ok i had a chance to try this one now too, it doesnt work either unless i did it wrong, you didnt show an example of what the changed lines should look like?

 

[wswd]

So there is still no update or fix on this...11 days later?

 

[cPanelMichael]

So there is still no update or fix on this...11 days later?

As of now, the plan is to push a new 11.42 build to the "Current" build tier tomorrow, then on to the "Release" tier on Wednesday, to address the proxy subdomain issues. A build that addresses these issues for 11.43 (11.44 development build) is scheduled for later today. Note that release dates can change due to unforeseen circumstances.

Thank you.