Search results

I tested the two virtual host configuration in crafted Cpanel style httpd.conf, both works. The problem might be somewhere else. Can you enabled DEBUG logging and test it again? To enabled debug logging, you need to change server "Log Level" to "DEBUG" and "Debug Level" to "high". You should...

No, there isn't. However, you can use Apache's httpd.conf, most commonly used directives are supported.

You need to create a context and set "Add Default Charset" and "Customized Default Charset". http://www.litespeedtech.com/docs/config/context/#addDefaultCharset

OK, I see, it is more like that setuid PHP process do not have read permission on php.ini. ;) For the other problem, clearing your browser cache may help.

That can be a problem, I am not sure Linux has this limit or not, will investigate. Using ACL is a good solution too, probably even better for system supports that. If phpinfo shows only "../conf" then PHP cannot found the php.ini in the give path, that path should be a relative path to the...

I think we will change the memory limit range, only >1M is allowed if it is not zero. ;) Please take a look "man 2 setrlimit".

Arkadius, It is not controled by ENV variables. you have to define a vhost level lsphp application and set "CGI setuid mode" to "docroot uid"

Has "63.247.93.106" been set to be a named virtual hosting IP?

That's for starting PHP from command line, no need for starting from within LSWS. No message probably becuse no system call within LSWS was failed, just the kernel kills successfully spawned process due to the memory limits. ;)

Yes, when an application is started by Litespeed. Yes, here should be. You can check the stderr.log and error.log, some error messages are hard to catch. :)

It is preferred to let LiteSpeed start FCGI applications no matter which Process Manager is used. I personally prefer LiteSpeed's built-in process manager. It is more intelligent, and have the following advantages 1) On-demand dynamic spawning, uses system resource more efficiently. 2) Fault...

I think your memory limit is way too low. Due to a bug in earlier release, application specific resource limit has not been enforced. So that might be the cause. And, if you let LSWS start the FCGI directly, the "-b" command line option is not necessary.

Hi Conundrum, Thank you for the tutorial. A few comments: 1) Initial Request Timeout is too short, should be set to something like "60", otherwise, litespeed may think the request has been failed for a request that needs more than 1 seconds to process. 2) Instead of starting the FastCGI...

CGI Chroot Mode should be "Same as Server" unless you want to build chroot jail environment for each every vhost. too much trouble. To let a vhost use this own lsphp, you should add a PHP LSAPI application with similar configuration as the global one. make sure the name is like...

120 templates! You kidding! :eek: If each template has only one vhost, you should just create the vhost directly. :) For a template, you should set template->security tab -> "CGI Resource Control" -> "CGI Set UID Mode". When you define an external app in a template, you should use some thing...

Arkadius, Are you talking about PHP suEXEC for multiple vhosts? You need to set http://www.litespeedtech.com/docs/config/vhostlist/#vhCGISetUidMode to "Docroot UID", then define lsphp external app and script handler for that vhost, cannot share the global PHP handler.

Answers to your questions: 1. It is possible but not recommended, as you need to build the jail environment manually for each every chroot vhost. that is not a easy job. Instead, we recommend chroot at server level along with suEXEC, our Enterprise edition has the best chroot support which can...

I agree, ETag generation should be application specific, otherwise, it is not that useful and wasting precious CPU cycles. :-)

And there is annother issue with generating ETag dynamically, web server have to wait till the last byte of response body before it can start sending response back to client.

Maybe we will add it when we implementing the disk cache. Calculating something like MD5 checksum for each every dynamic request is pretty expensive though. Any other idea for ETag calcuation? cgi_buffer uses MD5