Forking question

[CaptainChainsaw]

Hi all,

 

I'm looking at some code which has been rewritten so that the logic is forked.  I've created a test system and run the new code on my local machine, the code works properly.  I've been advised that the code was run live and it did work but the new code didn't perform as well as expected, there was nothing in the error log which would explain the issue.  As far as I know there's no limit to the number of processes that can be running concurrently but it does seem possible that the processes are being killed prematurely.  It's difficult to explain exactly what this code is doing.  The current live code processes data and the failure rate for example is 10%, moving to the forking process the failure rate increases.  Why it is increasing is what I'm trying to find out.  Possibly this is due to timeouts, db connections, process limits being reached etc, network issues.  The code seems to work but can't reproduce the increase in failure rate.

 

Anyone got any other ideas as to why the failure rate would increase when moving to forked code?

 

 

Cheers in advance,

 

CaptainChainsaw

[trq]

How could we possibly know without knowing the environment the code is being run in, or seeing any relevant code?

 

How exactly are you forking these processes?

[CaptainChainsaw]

Hi Thorpe,

 

I didn't want to post a pile of code without finding out exactly where the problem is as I didn't want to waste anyone's time.  I've found it that it seems to be a javascript problem so nothing to do with the forking as far as I can tell.

 

If I run into further problems with this I'll post something more specific.

 

 

Thanks again,

 

CaptainChainsaw