Experimental Physics and
Industrial Control System

That appears to be what is happening. For the area we had detecting it, I have been dropping those duplicates and that has fixed our issue. We may not have time to look into it much either, if we do figure it out then we’ll definitely let you know.

 

Thanks,

William

 

 

 

Hi William,

Here is the output for a few seconds.

Thanks; so the first monitor update is getting duplicated when it's coming from the CA gateway, and from Jure Varlec's post it sounds like this behaviour only occurs when running the gateway in its -no-cache mode. This confirms that the problem is internal to the gateway.

We don't have a dedicated maintainer for the gateway, and Ralph Lange who is nominally responsible for it doesn't have much if any time to work on it. Nobody is going to have any objections to your looking into the internals of the code and proposing a fix if you feel so inclined. In any case Ralph is busy all this week organizing the EPICS Collaboration Meeting that is going on at ITER in France.

If you are developing your own client software it should be relatively straightforward to detect and discard these duplications by comparing the monitor events timestamps, as long as you're using one of the data types which includes that information.

- Andrew

-- 

Complexity comes for free, Simplicity you have to work for.