I have a series of 7.6.11 Maintain forms that have been in production for years. We're transitioning to a new server and a new version of WF (7.7.03) so I have installed a new 7.7 Dev Studio and am deploying the files to the new server. They deploy, and they run, but when I click on a button in any of them I get

Application/Configuration Error
Failure, Node does not exist in odin.cfg

My Reporting Servers on the new setup have different names. But I don't find any references to the old Server names in any of the source code or compiled code. No EXEC calls, no CALL calls, no JavaScript that invokes WF . . . I'm confused that it will start up but won't continue to run.

The error message gives me very little to work with -- any suggestions on where to look?

Happy 4th by the way.

J.

Have you compared the odin.cfg files on the two servers to see if there are any differences - maybe in the database nodes specified therein?

Try looking for LOOPBACK in the server odin.cfg. I think that this is the one used by Maintain.

Should be identical attributes to the main server node, normally EDASERVE.

I'll check it again. I'm surprised it would run once but not twice.

J.

What I meant to say is I think LOOPBACK is used on a callback to Maintain. Maintain would start on EDASERVE or whatever through the client, but then makes calls direct back to server via LOOPBACK.

Mark?

My only advice is to make sure there are no typos or missing lines from ODIN.CFG. Also, some cases point to a file named ibicommd.wfs but that's not the same error that you are getting.

When the form is displayed, is the agent still active? We have seen cases when the form is displayed, the agent goes away. Then trying to perform ANY action will give an error.

Mark

The agent remains in use and I added loopback to my odin.cfg. That change made no difference in the behavior.

There's some complication here. This WF Client runs on port 7070 because the system has two WF Clients installed on it. If there's a hard-coded 8080 somewhere that could be the problem. There's also two separate Reporting Servers for 7070, one for QA and one for Dev. I coded the loopback to point to the Dev Reporting Server which is fine for now, but will be trouble when we put QA into the mix.

So to summarize, added loopback, flushed cache on Client, no change. Agent remains active and waiting for response.

As for syntax in odin.cfg . . . it pretty much looks correct, but I've never seen a published doc on it and the error message is vague as he1l. It also seems to run the first time through.

J.

John - Please open a case with the help desk so we can track and solve this.

Mark

I'll do that, but with a July 12th deadline I'm going to need to keep pressing on this one independently.

J.

Fixed for the moment. I needed to set a default Reporting Server for the Client. My concern is that with two different projects and each having a Dev and a QA Reporting Server that the default will only solve part of my problem. But setting that default to the server I'm running against right now solved my problem.

Thanks for tuning in gents.

J.