Php Process Killing itself
I have a website running on Joomla, PHP, Centos, WHM and on a specific part of it a script run's and does extraction of a zip file (the zip is already on the server).
However, the extraction never completes. The process ends with the command "+++ killed by SIGKILL +++".
Further research proved that SIGKILL is more of a linux term where the system is ending the process rather than a PHP timeout.
I have made sure that I have more than enough PHP limits. i.stack.imgur.com/1nYzJ.png
I have further noted down that the process ends exactly at 300 seconds every time. Is there a System level limit that needs to be changed here?
-
This is probably a KILL at kernel level caused by lack of resources eg mem + swap Have a look at Who "Killed" my process and why? and some of the excellent articles it links to, and see if anything there helps to make sense of the issue you are experiencing. 0 -
Hi, Thanks for the reply... I already did around 10+ days of google research and tried almost everything (to my knowledge) around that answer and others. The strange thing is that it always ends at 300 seconds and that seems to be a threshold. 0 -
If we start from the premise that the unzip command is failing with a resource allocation issue, perhaps you could do a test with a smaller zip file ? I would be looking at 2 possibilities: 1) The zip is so large that at 300 seconds run time, it has exhausted your server resources and is killed. 2) the zip is in some way corrupt and the unzip process is stalling and timing out at 300 seconds. 0 -
Hi, The smaller zip works but the program/script I am working with needs all the files in one zip. There must be a way to increase that 300 seconds limit? 0 -
This is probably the time to start talking to your server provider, and asking for anything in the logs that may help determine if you need more physical resources, or if it just needs some setting to be tweaked. 0 -
I have a ticket open with them and have been with them on this for about 8 - 10 days but they don't have a word back yet. 0 -
Ouch !! I know it's the holidays, but 8 days? You can probably guess what we would suggest you do ........... o_O 0 -
Ha ha... Well, they have been really really good all this time. This is probably the first time it's happening. I will be back with a solution 0 -
OK, the host got back to me and finally got it solved... It was the FcgidBusyTimeout value set to 300 in Apache config. 0 -
Excellent news that you have this issue sorted. Thank you for updating us with the solution (and I have learned something new as well :-D ) 0 -
Well, Thank you for guiding through Grumpy Old Man, some credit goes to you sir. 0 -
Thanks for updating the thread with your solution. :) 0
Please sign in to leave a comment.
Comments
12 comments