@jasahu

We ended up putting in a “solution” where when IFS would throw an error that it “Couldn’t find Quickreport….” (paraphrasing that) we would catch that error and then call a custom action that would call Dictionary_Sys.Refresh_Odata_Projection_Cache(“QuickReports”, “”).  Alerting the user that whatever they were trying to do failed and to try again.  Most of the time, the refresh would be done by the time the user tried it again and wouldn’t see the error again.  Or the refresh would still be happening and then once finished it would be there.

We have updated IFS a couple of times since this issue, but I can’t say with any confidence that it would be resolved.  We haven’t had the issue since putting in this solution.  But it also may never be happening anymore either.

Hi @hansend,

Thanks for detailed info, that is quite a trick.

Just be sure if I understand it well:

Does the error occur without you touching the definition of quick reports involved?

You just keep using the solution as intended without changing it and sometimes suddenly this error “Couldn’t find Quickreport….” appears?

@jasahu

From our experiences, if any Quick Report would get touched it would refresh the cache, which made sense.  Most of the time the refresh didn’t take a lot of time, came back and the user didn’t even know the wiser.  Just the occasional refresh that dropped certain Quick Reports.  We also adjusted when we updated and/or installed Quick Reports into production to prevent fewer cache refreshes.

And yes, we just created a service that we run all our Quick Reports like this through.  Since implementing, I haven’t really looked through our logs to see if it is still happening.  I also haven’t received any support calls for it after we trained our users over it and explained to them what is going on.  

One odd thing that we discovered is that this almost always happened to us when we updated IFS and even applied the occasional patch.  So we made sure in our scripts we called that API when it was all done.

Thanks, this is really good to know.
One thing I am still trying to find out:

When this whole problem started, at the beginning with sporadic errors, did you get the error without touching the QRs, or this error only appeared when you created/modified/removed QRs?

In our case it would be used for machine-to-machine communication and we might not be able to use your approach with that intelligent exception handling+cache refresh+retrying.

Before you implemented this exception handling, did this error sometimes come without modification of QRs, just by normally using the interface?

@jasahu 

They only disappeared when creating/modifying/removing QRs and then on the occasional upgrade.

So if it was there, it always stayed there.