Hello,
Lately I have noticed a client's Evo websites (6 on the same shared hosting on a reseller account) displaying internal server error messages in one or more of the Evo manager frames when logging into, or generally using, the Evo Manager.
The internal server error message can usually be removed by refreshing the frame. When the issue is happening, it always appears on login. It also can appear after saving a resource or element.
The error message usually occurs in only one of the manager frames at a time, with the "main RHS" frame being most common - but it can also appear in the "LHS tree" frame, and less occasionally in the "top header" frame.
I've noticed this happening recently on some client sites (all on shared servers) at a couple of different web hosts.
It started happening on Evo 1.0.12 (now 1.0.14) sites on the same reseller account after a short period of time (say a month or so) of being moved to newer shared hosting servers which has "Cloud Linux" installed (case A).
* MySQL 5.5.25-cll
* PHP 5.3.28
* memory_limit 256MB
The issue never previously occurred on these sites on the older shared hosting servers at the same host.
* MySQL 5.0.96-community
* PHP 5.2.17
* memory_limit 64MB
It also happens on a different Evo website (1.0.10) (case B) which is at a different webhost. Although interestingly, when I logged into this Evo manager to today to check version numbers the error did not occur which is the first time ever.
* MySQL 5.5.36-cll
* PHP 5.3.27
* memory_limit 64MB
So I am not sure if the error is related to Cloud Linux - or a setting on the server - or something in Evo which causes the problem.
(I have never seen the issue on WordPress or PrestaShop sites.)
After going back and forth with one of the Australian web hosts, it was suggested that its to do with resource usage. On the main lot of servers that I am talking about above (case A), a 1 GB memory limit applies (which I believe refers to RAM). The host says "This 1 GB memory limit is assigned to each cPanel user on the server. If the websites hosted under that user exceeds that limit then only website shows the internal server error. If website run within that limit then website will not show this error anymore."
When the host resets the "Cloud Linux", the case A problem goes away and all manager frames load as expected - no error messages appear. Unfortunately we cannot control the resetting - that is handled by the host and applies to all sites on the server so is not a viable solution.
One of the case A sites is medium size (100 pages) and gets a fair bit of traffic. The others are just small sites with standard traffic. The case B site is small in size and traffic.
The case A client contacted the host about the issue directly today and was advised something different: "For that you must contact your developer and ask him to optimize the website's PHP memory_limit usage". How can I do that? And regardless, the memory_limit for the case A sites is bigger than the older server than had no issue.
I'd love to know what is causing the issue and wonder if any other Evo users have experienced the internal server error message in one or more of the Evo manager frames - and what can be done to stop if from happening again?
Thanks in advance.
[ed. note: nickf08 last edited this post 12 years, 3 months ago.]
me too , I have a same problem.
My old hosting -
Apache version 2.2.24
PHP version 5.3.27
MySQL version 5.5.36-cll
My new Hosting -
Apache version 2.2.24
PHP version 5.3.25
MySQL version 5.5.35-cll
Thank you .
3 thoughts...
1.
http://wiki.modxcms.com/index.php/What_is_mod_security_and_how_does_it_affect_me
2. File permissions 755/644
3. You normally set memory_limit in a php.ini file but can be done through htaccess
Quote from: nickf08 at Jun 18, 2014, 10:25 PMWhat is the recommended memory_limit for Evo 1.0.14?
I'm using 128MB without problems but if you use 3rd party apps such as ecommerce etc. go for 256MB or more.
Quote from: zawlinoo at Jun 20, 2014, 03:47 AM
And Error log :
[Fri Jun 20 01:41:05 2014] [error] [client xx.xx.xx.xx] (12)Cannot allocate memory: couldn't create child process: /opt/suphp/sbin/suphp for /home/msc/public_html/mydomain.com/index.php, referer: http://www.mydomain.com/manager/index.php?a=1&f=menu
Hmm...
https://www.google.com/search?q=/opt/suphp/sbin/suphp
Ask you hosting company if they can see what is different in your setup between the other sites working.
I believe traffic has to do with it and a VPS would help. Why not ask the hosting company if you can try it for free
a month.