Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [ecf-dev] Problems with IRemoteServiceContainerAdapter

Hi Scott,


Ok....I believe I see the bug...in your use case the call to

      synchronized (localRegistry) {
addReferencesFromRegistry(clazz, remoteFilter, localRegistry, references);
      }

actually adds a remote reference to the references List (because the localRegistry *also* publishes the service). This means that the following test

if (references.size() == 0) fails, and the associated block (which sends the AddRegistrationRequest) is not executed, and the remote reference is never found.

Yes, that's exactly the problem. Thanks for fixing it.

Regards,
Eugen

Am May 14, 2009 um 21:39  schrieb Scott Lewis:

Hi Eugen,

Eugen Reiswich wrote:
Hi Scott,

The code in red is for getting local access to published remote service references...it shouldn't block (but apparently is in this case...or is it failing outright?).

Well it doesn't really block anything but once this code is executed it does search for local remote services and finds some (because we have registred them as remote services) but then (that's what we guess) the connection between service provider and service consumer is never established as the AddRegistrationRequests are never sent. Again, this is just what we could observe. When we comment out this part everything works fine.

Ok....I believe I see the bug...in your use case the call to

      synchronized (localRegistry) {
addReferencesFromRegistry(clazz, remoteFilter, localRegistry, references);
      }

actually adds a remote reference to the references List (because the localRegistry *also* publishes the service). This means that the following test

if (references.size() == 0) fails, and the associated block (which sends the AddRegistrationRequest) is not executed, and the remote reference is never found.

I've opened critical bug https://bugs.eclipse.org/bugs/show_bug.cgi?id=276368

I expect to have a fix/patch available for review within a few hours (the fix should be/is easy) and after review by other ECF committers it will be released to HEAD.

Thanks Eugen for the report. Once the fix/patch is in place for review, I would appreciate it if you could test it out in your environment/use case to verify that it deals with your issue.

Scott


_______________________________________________
ecf-dev mailing list
ecf-dev@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/ecf-dev



Back to the top