Cron problems, empty REMOTE_ADDR, random freezes

[Himeko]

Hi,

Opened a ticket about this a few hours ago but it's still unreplied, maybe someone can give me a hint how I can solve this myself.

I joined Jaguar ~3 days ago and finished moving my sites over from my other server yesterday. I tried to reinstall everything as close as possible and matching the settings, same CentOS architecture, same php and mysql versions, same apache modules, then went over the settings extensively on both servers to match the configuration.

Yet I have a problem here with crons, which ends up with the scripts unable to run. Checking the cron output that's being sent to my email I get various different responses depending on the method to call the cron. I tried with lynx/source, php -q, wget, both with URI or /home/user.. paths to the scripts, direct calls, none work. I get errors like "Could not open input file:", "/bin/sh" + "No such file or directory", permission denied..

Also another issue with my ticketing software. It has a cron to fetch emails and parse for tickets, which obviously isnt being called either, and if I call it from ssh a very weird issue occurs:

It has a function that checks the caller IP against a list of allowed IPs; so it has by default 127.0.0.1/localhost, and it checks it via the $_SERVER['REMOTE_ADDR'] variable. The problem is, when called from the console (with php .. script), this variable is empty. Upon further testing, I dumped the entire $_SERVER array, and the REMOTE_ADDR variable isn't even there! If I call the script via the browser, the address is there, just not when called from inside.

Any ideas how to fix these issues?

Also I've noticed, every few hours my SSH console gets disconnected. I didn't think much about it the first time, even the second, but now I'm noticing this as if a pattern. When the SSH gets disconnected, nothing else works for a few seconds .. 5..10.. email doesnt connect, apache not responding, even though the vps is still up. Not a major issue but as a developer quite annoying when I'm editing files straight from the server, monitoring, etc, it all goes to waste (for example files opened via WinSCP, they aren't "linked" anymore after reconnecting and I need to re-download/reuplaod changes etc).
Is this an issue in my VPS or something specific to the Jaguar VPSs? Maybe some service that runs every x hours to do something on the nodes or?

Thanks.

[Ron]

I have been searching the forums for the remote_addr issue... I thought we had a problem like that when we switched to suPHP or just before then. I can't recall the exact details.

[JPC-Anoop]

Hello Himeko,

I have replied your ticket. Please check it and update the ticket to follow up on this.

[Himeko]

Please check back on ticket 13283275, php isn't working now at all, Error 500 on all php sites for over half an hour now.

[Himeko]

I can't wait any longer, customers calling for over an hour with their sites down and I have no reply whatsoever to my tickets. I'm rebooting the server and rebuilding apache/php to try and get it working again, nvm the crons issue, this is very serious when all my customers sites are down for such prolonged time.

[Himeko]

The rebuild fixed the issue with sites not loading. I'll look into the crons problem myself tonight.

[JPC-Stefan]

We are sorry for the delay in response. Your tickets have been updated.