Configure time limits for a job (and alert/cancel if beyond)

1. there won't be a distinction between a manually cancelled job and a job cancelled this way or any other at present. If you feel this is important, could you add a new entry for this idea?  https://knowledge.safe.com/content/idea/post.html?space=153  2. This time will not include queue time. Only running time. 

I've created two new entries for each of your ideas @ngoorman. I think that others may want these as well. So, I'm hopeful that we'll see the votes pile on! Please do add your comments to them   https://knowledge.safe.com/idea/38533/configure-a-global-system-wide-running-job-expiry.html  https://knowledge.safe.com/idea/38534/configure-a-set-running-job-expiry-time-for-each-w.html  Thanks for contributing!  

At the moment we are not a notification is not being provided. If you feel this is important, could you add a new entry for this idea?  https://knowledge.safe.com/content/idea/post.html?space=153 I appreciate your help making FME even better!   

It sounds like the feature we added will help you out :-)  In regards to the memory limit, we don't currently have plans to add this feature. If you feel this is important, could you add a new entry for this idea?  https://knowledge.safe.com/content/idea/post.html?space=153  Thanks for helping out with FME and ideas!

Please see additional ideas here:  https://knowledge.safe.com/idea/38533/configure-a-global-system-wide-running-job-expiry.html https://knowledge.safe.com/idea/38534/configure-a-set-running-job-expiry-time-for-each-w.html

I've created an idea for the system wide defaults, which is not an option at present. Please do vote for it and comment. Some good news for is that the new running job expiry time parameter is available in the job scheduler. '  Thanks for contributing!

+1 from me, especially for users with a single engine this would prevent one process from blocking it for too long. An option to include a message back to the user would be nice too.

In FME Server 2017.0 beta 17160, you can configure a job to cancel after a period of time using the REST API or a transformation service. Time Til Canceled (ttc) is the configuration parameter to set. You specify the time in seconds and the minimum value is 60 seconds.'REST API'Transformation Servicehttp://ap-fme32bit/fmejobsubmitter/akoning/forever.fmw?opt_showresult=false&opt_servicemode=async&tm_ttc=60

Aaron, Would the same/similar functionality be available as system wide defaults (max threshold of like 12 hours) and/or available to implement when scheduling workspaces? We use FME Server Scheduler to run almost all the workspaces that we have on FME server, so if it's solely in the REST API, it probably wouldn't be of much use.  

Yes please! :) The need for such an option was also discussed here by the way. An option to cancel a job if it takes up more than x GB of memory would also be on my list of wanna-haves actually.

Hi @aaronkoning! Thank you (and the Safe team) once again for implementing this! Two questions I have though: When the job is cancelled, will I be able to tell that it was aborted because it timed out? Or will the job status just say CANCELLED? If the latter, I would love to see an additional status, e.g. TIMED_OUT... I assume this time limit does not include the queue time, right? So the time-out only applies to actual engine processing time?

This is a very annoying kind of a problem. We have severe problems with hanging WFS reader jobs, blocking engines and the whole processing chain! We need a solution here immediately! See also: https://knowledge.safe.com/questions/22981/wfs-rea...

What @aaronkoning mentioned basically is a time limit per workspace, if you realize that 1 job == 1 workspace. Please note that we are talking about FME Server (not Desktop) in this topic: you submit a job to run a workspace using some parameters (in this case posted using the REST API). Also, if you already have such a nice fine-grained solution, why would you still need a global time limit? Although there might be a hidden setting for this in the Job Routing section of the server config. Anyway, thanks for implementing this, Safe! 

A good start, but I'd very much like to have the ability to set a global time limit, and preferably also a time limit per workspace.   

This has been released as part of FME 2017.0 . Visit http://www.safe.com/downloads to download and give it a spin.