06-04-2012 01:05 PM
Ok... that's easy enough. What hardware do you have exactly? It might be specific to one CRio type or systemic.
06-05-2012 02:40 PM
FYI, I have heard that there is a small memory leak associated with using LVOOP on cRIO. As I recall the leak is due to LV not recovering all of an object's memory after it is no longer needed. It is apparently a very small leak and won't be an issue unless you are creating a lot of temporary objects--which you do in a messaging system--and your system will be running for a long time.
Has anyone else heard about this or is able to quantify the leak?
06-05-2012 03:46 PM
I have not heard of any such behavior. I do know that CRios do not do memory compaction. Is this what you are referring to?
06-06-2012 11:30 AM
psyklopz wrote:
Whoops, I was wrong, the following doesn't work... I got back from lunch and nothing worked anymore.
To clarify, did using the 2010 back port work?
I tested the VIs you sent me and I can successfully run them on a cRIO 9024 but they don't seem to work on the 907x cRIO series. The VIs deploy but then connection is lost as if the target crashes before the top level VI begins to execute. Is this the same behavior you were seeing? We believe we've fixed the issue in 2012 but I want to clarify whether the 2010 back port is a viable workaround for you for now.
06-06-2012 11:55 AM
The 2010 back port does not work for us. Every once in a while it would run, but would crash about 95% of the time right on launch, similar to the 2011 framework, and the behavior you were experiencing.
06-07-2012 10:27 AM
psyklopz: Try this... go back to the 2011 version of the framework, but use this VI instead of the Launch Actor.vi that is in the framework...
https://decibel.ni.com/content/docs/DOC-22754
The crash seems to come from the static VI reference, so I have written this VI to eliminate the need for the static VI reference. I have not tested this on CRio, just on my desktop.
06-07-2012 11:52 AM
AristosQueue wrote:
I have not heard of any such behavior. I do know that CRios do not do memory compaction. Is this what you are referring to?
It was definitely explained to me as a memory leak, though it is possible something was lost in the translation from NI -> Other Developer -> Me. I'll send you more information via PM you might be able to use to track down the source and verify/clarify the issue.
06-11-2012 01:11 PM
This doesn't seem to work. We've spent alot of time working with it to no avail. We can't get a few of our more complex actors to even deploy to the cRIO. It seems to be a problem with having multiple actors, or actors which launch other actors. I'm not sure which.
06-14-2012 06:38 PM
It appears to be a problem with Community Scope VIs not deploying to the CRios. Go into the Actor Framework and make all the VIs that are Community Scope be Public.
06-14-2012 06:51 PM
Will that be fixed (somehow) with LV 2012 + AF 2012? I don't imagine we'll be advised to change the scope of our vi.lib VIs when it ships...that change wouldn't propagate to other dev environments, anyway.
I'd like to know whether deploying AF applications to cRIO/sbRIO will be production-ready in LV 2012.