508 Errors?

[optize]

We're seeing a lot of complaints about 503 and 508 errors now with 4.2.5. Most of the time, it's due to hitting CloudLinux limits, so I'm trying to figure out if this is a recent change in Litespeed or CloudLinux.

It seems the complaints started happening as soon as we upgraded to 4.2.5.

 

[wanah]

4.2.5 is supposed to make Litespeed compatible with Cloudlinux. Before when Cloudlinux produced a 508 error litespeed didn't understand it and produced a 503 error and restarted.

Now a 508 error is shown and litespeed isn't restarted. You shouldn't see more errors but should have fewer restarts. If you are seeing more erros then maybe they were either hidden by listespeed's reboots or maybe there is something else wrong.

You shouldn't see any 503 errors only 508 errors on sites that need either more ressources then you have allowed them or some optimisation or some form of anti ddos if they are under attack.

 

[optize]

Has anyone reported slowness when there's a lot of 508 errors?

I had a situation last night where the load on the box was nothing, IO was nothing, the box was sleeping, but all web requests to the box took forever (15 seconds+) to load simple pages.

I saw there was someone getting a lot of traffic, which was all being throttled by CloudLinux, so you were spitting out a ton of 508 errors.

I ended up downgrading to 4.2.4, and the issue went away. (The user was still being throttled, but it didn't affect other customers).

I don't have time to troubleshoot the issue with you, but figured to let you know.

 

[mistwang]

Which PHP suEXEC mode is used? regular, through CGID daemon, suEXEC daemon mode?
Do you have error log and stderr.log when the server is using 4.2.5, please send it bug@litespeed....