Server Fail

[nickbits]

Hi All,

Looking for suggestions here after getting no where with EUKHost Ticket Support.

For the past 3 days I have had my server, or at the least the web service falling over. Today (2 Apr) I had a monitor warning that all my sites went offline at 5am for at least 5 minutes. The same occurred for at least the previous two days, but at different times. One was between 9-10pm.

So question is why? Nothing has changed that I can see. Am assuming it is something to do with this next bit:

in lfd.log I am getting lots of warnings/errors like this:

Apr 2 09:22:49 server lfd[30272]: *User Processing* PID:30267 Kill:0 User:***** VM:194(MB) EXE:/usr/bin/php CMD:/usr/bin/php /home/******/public_html/index.php

Now I am using Drupal, I know the files are fine, I do not get any issues with it on my development machine and have not had any issues until a few days ago. I do not think anything has changed, infact on one of the sites I know nothing has changed.

So the question is what is causing this and how do I track it down?

I would appreciate any pointers?
Regards,
Nick

[nickbits]

Today (3rd Apr) I have had another server fail (httpd/apache) at 9.35am. I am looking through the logs, but not getting too far.

I am loosing patience here. I have tried Livechat, and get no where, thats usually after waiting 10mins or so for someone to answer. Not only that, sometimes I have been left waiting 5 or 10 minutes even after I get through to someone. And then the answer is usually totally rubbish, something like:

We have experienced a high load...

or

It is your script causing the problem.

Well my sites, some have been updated, some have been the same for months, but the problem only started 5 or 6 days ago so I know it is not the scripts.

The last ticket I submitted (FMU-69538-989), I am still waiting for a reply to.

I am loosing patience and confidence in EUKHost and am looking at alternatives.

[Kieran]

Hi Nick.

I'm looking into it. I will check your ticket and server status and update you soon.

Regards,
Kieran. A.

[Kieran]

Hi Nick.

I have referred all the replies from ticket FMU-69538-989 and also checked your server configuration. As per our investigation and administrator's replies, your server was overloaded due to one of the account and then affected with memory issue.

The apache service was killed due to OOM ( out of memory) so we have temporarily increased memory on your VPS. We have recommended you to keep an eye on it and let us know if you are still facing issue related to services.

Please update us in same ticket if you have any query in this regards.

Thank You.

[nickbits]

Thanks for that. Will reply in ticket, but what is funny is that the problem has only been occurring in the last week. So I would prefer if we can to find the cause.

[nickbits]

Issues are back. I have updated my ticket but thought I would reply here as well. For the past 3 days the sites on my server have ran fine. Then today the server (httpd) has crashed twice. Like I said before, it has been working fine until last week when all the problems started. Nothin has changed, nothing major and I am getting annoyed with my server keep going down, I need to know what has changed to cause these problems.

[nickbits]

After more tickets and live chats, I woke this morning (8th Apr) to a reply to say issues were sorted. 5 mins after the email and ALL of my sites are down.

Okay, problem now is none of my sites work, all coming up with "page not found"

So EUKHost, whatever you have done has messed everything up.

It looks to me like you have disabled APC and re-enabled Zend. Don't know why as APC was working fine.

Can you get this sorted please, I am growing fed up of these problems now, two/three weeks to get it sorted is more than enough time.

I did leave this Live Cha a few mins ago, apart from being SLOW, nothing new was noted...

I am getting to the critical point now, three weeks of issues, nothing has been resolved. I have been a customer for 3 or 4 years with no major complaints, it is only the last few weeks I have had serious issues. I don't want to change providers...

[eUK-Alan]

After more tickets and live chats, I woke this morning (8th Apr) to a reply to say issues were sorted. 5 mins after the email and ALL of my sites are down.

Okay, problem now is none of my sites work, all coming up with "page not found"

So EUKHost, whatever you have done has messed everything up.

It looks to me like you have disabled APC and re-enabled Zend. Don't know why as APC was working fine.

Can you get this sorted please, I am growing fed up of these problems now, two/three weeks to get it sorted is more than enough time.

I did leave this Live Cha a few mins ago, apart from being SLOW, nothing new was noted...

I am getting to the critical point now, three weeks of issues, nothing has been resolved. I have been a customer for 3 or 4 years with no major complaints, it is only the last few weeks I have had serious issues. I don't want to change providers...

Hi Nick,

This has already been resolved as explained in ticket FMU-69538-989. Disabling apc fixed the problem with your sites.

Please feel free to update the ticket in case you need further assistance.

[nickbits]

Hi Nick,

This has already been resolved as explained in ticket FMU-69538-989. Disabling apc fixed the problem with your sites.

Please feel free to update the ticket in case you need further assistance.

A few things:

1. The problem has not been resolved, the server crashed again today (13th at 3.30PM)

2. APC was not the issue, I ahve been running APC fine since January. As I said in my ticket, the settings on the server have been changed, ZEND optimizer was turned back on which is known not to work with APC. However I ahve not turned ZEND back on and wanted to know who did and why.

3. The ticket system also now doesn't work!

All of this is adding to my frustration with the system at present. As I said before I have been with EUKHost for 3 or 4 years. Until a few weeks ago I have not had any major problems or reasons to complain. The problems occuring are not down to anything I have done, infact some of the sites affected have not been altered for 3 or 4 months so something has changed that I do not know about. I can no longer stand for servers crashing...

[nickbits]

So I cannot create a ticket since the new support system doesn't work! I emailed support and have eyt to get any reply. And no reply via the forums. Come on guys this is rediculous, how are we to get support?

[MattBarthorp]

I am sorry to hear about the problems you have encountered. Being a customer of Eukhost and based on the experience so far, I strongly believe that the issues related to your server should soon get sorted. Yes there has been certain minor issues with the ticketing system since the maintenance. Yet, I believe they are making some improvements inorder to make communication better.
I would suggest you to get in touch with them via. Live Chat available on their official site.

[nickbits]

Please try other CMS. such as joomla. Perhaps it will work

I don't want to use them, I love Drupal and it is not Drupal that is the problem. Besides Drupal is far better than Joomla (in my opinion), but I do realise everyone has their own fav.

I have been running Drupal for over 4 years and have only had the problems in the last few weeks. As I have mentioned before, not all the sites that are affected have been touched for several months and not all use Drupal. Hence it is not the code that is at fault as far as I am concerned, unless it has been changed without my knowledge, but apart from EUKHost, I am the only one with access.

[nickbits]

Have jsut had a reply via my ticket which is totally useless...

We have disabled Zend on your VPS which may conflict with APC.

In the first case EUKHost told me it was APC causing the problems and hence disabled it. Now Zend has been disabled but has APC been turned back on?

Regarding downtime, as we mentioned in previous ticket, your VPS is exceeding memory which is clearly showing in following logs collected from the main hardware node,

First off I cannot see my previous tickets due to the fantastic new system that has been put in place!

As for the memory issues and server crash, I already know that. What I want to know is how you are going to fix it. You have told me several times you have fixed it. You have told me:

1. You disable APC because that was the issue
2. You disable lfd because that was the issue
3. You increased memory because that was the issue.

I am sorry but 3 solutions and none have done the job. I am seriously getting to the end of my rope with this. The issue has been going on for about 3 weeks now, you keep telling me it is fixed, but clearly it is not.

Please can you sort it out, SOON.

[nickbits]

I would suggest you to get in touch with them via. Live Chat available on their official site.

Thanks for the comments. However, firstly I ahve largely given up with Live Chat. The last 2 or three times I ahve had to wait more than 10minutes for someone to come online. Then I am left waiting for a reply. Then all I keep getting is fobbed of. For example one of the previous times I talked to them was on the issue I have talked about here. I mentioned to them the server had failed, all the details, etc. So the guy reset the services running. I just could not make them understand the problem. Another one said it was something I had done...

As for getting sorted soon, this is now 3 weeks old, so define "SOON"...

Yes there has been certain minor issues with the ticketing system since the maintenance. Yet, I believe they are making some improvements inorder to make communication better.

Minor issues, I am unable to login to the ticket system full stop. How can that be making communication better. Not only that, other threads have also indicated we will loose all our existing tickets!

Sorry, not getting at you, I am just frustrated at EUKHost. over 24 hours since I raised the issue, I ahve had one update and that contradicts a previous update. So you can see why I am growing frustrated. Again, like I said before, have been with EUK for 3 or 4 years and this is the first major issue I have had.

[nickbits]

So 48 hours after last requesting the issue get fixed and all that has happened is ZEND has been turned off and APC apparently turned back on. Still no solution to the problem or GOOD reason why/what is causing it.

[nickbits]

We are now 3 days since I started getting problems again and 3 weeks since they first showed. I have had no help this week at all, other than to turn APC off, Zend on and then Zend off and APC on. I am being given total rubbish answers, no real support and contradictory reasons on what is causing the problems..

I have had no update via forums or tickets (which I still cannot login to). My server went off line at 1.30 this morning. You have had over 3 weeks to resolve the server crashing, more than enough time.

Enough is enough, get it sorted.

[nickbits]

Okay 4 days since I re-raised this now 3/4 week old problem, and still it is not resolved.

[nickbits]

Hi Nick,

This has already been resolved as explained in ticket FMU-69538-989. Disabling apc fixed the problem with your sites.

Please feel free to update the ticket in case you need further assistance.

The problem has not been resolved and I cannot update the ticket as the fantastic new ticket system will not let me log in.

[Rock]

I don't want to use them, I love Drupal and it is not Drupal that is the problem. Besides Drupal is far better than Joomla (in my opinion), but I do realise everyone has their own fav.

I have been running Drupal for over 4 years and have only had the problems in the last few weeks. As I have mentioned before, not all the sites that are affected have been touched for several months and not all use Drupal. Hence it is not the code that is at fault as far as I am concerned, unless it has been changed without my knowledge, but apart from EUKHost, I am the only one with access.

Have jsut had a reply via my ticket which is totally useless...

In the first case EUKHost told me it was APC causing the problems and hence disabled it. Now Zend has been disabled but has APC been turned back on?

First off I cannot see my previous tickets due to the fantastic new system that has been put in place!

As for the memory issues and server crash, I already know that. What I want to know is how you are going to fix it. You have told me several times you have fixed it. You have told me:

1. You disable APC because that was the issue
2. You disable lfd because that was the issue
3. You increased memory because that was the issue.

I am sorry but 3 solutions and none have done the job. I am seriously getting to the end of my rope with this. The issue has been going on for about 3 weeks now, you keep telling me it is fixed, but clearly it is not.

Please can you sort it out, SOON.

Thanks for the comments. However, firstly I ahve largely given up with Live Chat. The last 2 or three times I ahve had to wait more than 10minutes for someone to come online. Then I am left waiting for a reply. Then all I keep getting is fobbed of. For example one of the previous times I talked to them was on the issue I have talked about here. I mentioned to them the server had failed, all the details, etc. So the guy reset the services running. I just could not make them understand the problem. Another one said it was something I had done...

As for getting sorted soon, this is now 3 weeks old, so define "SOON"...

Minor issues, I am unable to login to the ticket system full stop. How can that be making communication better. Not only that, other threads have also indicated we will loose all our existing tickets!

Sorry, not getting at you, I am just frustrated at EUKHost. over 24 hours since I raised the issue, I ahve had one update and that contradicts a previous update. So you can see why I am growing frustrated. Again, like I said before, have been with EUK for 3 or 4 years and this is the first major issue I have had.

So 48 hours after last requesting the issue get fixed and all that has happened is ZEND has been turned off and APC apparently turned back on. Still no solution to the problem or GOOD reason why/what is causing it.

We are now 3 days since I started getting problems again and 3 weeks since they first showed. I have had no help this week at all, other than to turn APC off, Zend on and then Zend off and APC on. I am being given total rubbish answers, no real support and contradictory reasons on what is causing the problems..

I have had no update via forums or tickets (which I still cannot login to). My server went off line at 1.30 this morning. You have had over 3 weeks to resolve the server crashing, more than enough time.

Enough is enough, get it sorted.

Okay 4 days since I re-raised this now 3/4 week old problem, and still it is not resolved.

The problem has not been resolved and I cannot update the ticket as the fantastic new ticket system will not let me log in.

Hi Nick,

It has been stated several times in FMU-69538-989 that your existing VPS (L-VPS-01 which has a Guaranteed RAM of 512MB) is facing memory shortage issues, hence we temporarily added additional 512MB more RAM (total 1024MB) to overcome this issue. We did provide you the kernel logs which clearly show that the processes are being killed due to Out Of Memory (OOM) errors. I'd recommend upgrading the VPS to a higher plan or optimize the scripts/applications which are responsible for hogging the memory, either will help fixing the problem. I've now again set the memory limit back to what it was according to the package (512MB).

We do have a 24x7 Live chat & Phone/call support available through which you can contact us regarding these mentioned issues. Again, about the helpdesk login not working, we can get this sorted out, only if you contact us from a proper channel rather than posting on forums.

Please let me know if you have any further questions on this.

[nickbits]

Hi Nick,

It has been stated several times in FMU-69538-989 that your existing VPS (L-VPS-01 which has a Guaranteed RAM of 512MB) is facing memory shortage issues, hence we temporarily added additional 512MB more RAM (total 1024MB) to overcome this issue. We did provide you the kernel logs which clearly show that the processes are being killed due to Out Of Memory (OOM) errors. I'd recommend upgrading the VPS to a higher plan or optimize the scripts/applications which are responsible for hogging the memory, either will help fixing the problem. I've now again set the memory limit back to what it was according to the package (512MB).

We do have a 24x7 Live chat & Phone/call support available through which you can contact us regarding these mentioned issues. Again, about the helpdesk login not working, we can get this sorted out, only if you contact us from a proper channel rather than posting on forums.

Please let me know if you have any further questions on this.

Hi,

It has not been clearly stated that memory was doubled, it was only increased to 768MB (or around that). That was my understanding. Anyway I seem to be stating the same thing again and again. Firstly I am well aware that I am getting the OOM errors. The point I am trying to make is that it did not start occurring until 4 weeks ago and at that time nothing had been changed. So once again I am asking what has changed to cause these errors?

Even with the increased memory the problems were still occurring, so I don't see how changing the package will make a difference, there is obviously a more fundamental issue. I am not willing to change the plan when the errors have occurred overnight and I have not changed anything.

As for the tickets and proper channels, I have been on live chat and raised it over email, that is why I have been posting on the forums because I don't seem to be getting anywhere.