Search results

Thanks to a reliable bug report, we fixed a 100% cpu bug which happens after the connection limit has been reached once. It can be easily reproduced by set "Max SSL connections" to "0", then access a https web site, then access a normal web site. Fix is in up coming 4.1.8 release.

Do you still get recv_fd() failed error in the log? Does the problem starts immediately after upgrade or wait for a while? Are you using caccti integration to monitor the server?

A record for www.

Figured it out, it is due to Apache worker MPM is used. edit /etc/sysconfig/httpd comment out HTTPD=/usr/sbin/httpd.worker it works properly now.

tested adding a new subscription in our plesk 10 test server, also works fine.

We tested plesk 10.3 on centos 5. Maybe we need to take a look at your server if you do not mind send us a temp root access. you can create ticket under bug report department.

We installed plesk 10 in our lab. Seems LSWS with using httpd wrapper is working fine. We tried "restart" apache from service management page of plesk, LSWS restarts, we added a new account, LSWS restarts as well. Please tell us more about your problem, what kind of operation you did and LSWS...

If you switch back to Apache, does PHP/Mysql work properly? if yes, you need to rebuild lsphp5 from "build matching PHP".

The global WAN speed should be very fast those days, you can give it a try from your location by update "/etc/hosts" by pointing your domain to our proxy server, see if the speed is good. I think it should be better than being taken down by botnet even if it was indeed slightly slower.

Our anti-DDoS proxy is up now, you can sign up at https://store.litespeedtech.com/antiddos/cart.php

Please compare phpinfo() page from LiteSpeed and Apache, PHP configuration should be similar, with similar PHP modules. If you use suphp with Apache, you need to turn on PHP suEXEC on LiteSpeed, if not, you should disable PHP suEXEC.

You should get enterprise edit if you worry about ddos attacks. the standard edition cannot handle large botnet.

Please PM me your server IP. we need to investigate. Your downtime is likely caused by something else, please send us the whole error log for analysis, failure to contact update.litespeedtech.com should not bring down LSWS.

We have a solution to detect if the request is sent by botnet or a real user, if detected botnet, our service will block those IPs at firewall level. I recommend you sign up with our anti-ddos proxy service when it become available (probably this coming Monday). It is free during our trial period.

Is the attack a GET flood? Can you tell which URL the botnet abuse? For large scale attack, the built-in anti-DDoS may not able to stop, it is depends on the size of the botnet and how aggressive the robot behave. Our antiDDoS proxy service will be live soon, maybe you can give it a try.

That's normal for PHP suEXEC, it will change user ID and execute lsphp5, usually it is not visible in ps, nothing to worry about.

Our lscgid should be listed as in "ps -ef" output, and usually only one. not httpd -n ##

It is not related to LSWS. You should find out what it is doing with "lsof" and "strace".

Put it in httpd.conf, or modsec2.conf modsec2.user.conf, just like your other modsec rules.

use pstree to check who own those processes.