DESCRIPTION

JCBC-229 : Tests added for memcached connection, the vbucket config and the Resubscriber Back off functionality.
           The last test caters to the changes made in JCBC-227.

Change-Id: I8309a39fa503402e996e413b3f3d3f48f44f0d0d


COMMENTS

author: Yasmin Rush
date: 2013-02-14 10:52:36.201000000

Uploaded patch set 6.

-------------------------------------
author: Jasmin Rangel
date: 2013-02-14 10:52:38.194000000

Patch Set 6:

New patchset patch-id matches previous patchset, but commit message has changed.

-------------------------------------
author: Natasha Sullivan
date: 2013-02-14 14:20:53.919000000

Patch Set 6:

That's much better, but still not correct I think. Because you're not testing the backoff actually. Backoff in this case means that the do/while loop always pauses longer until the next request comes in. So in your case (since I guess all your hosts are not accessible, it will loop indifinetly!) Your mock needs to store the calls it gets and measure the time difference. You can then add methods to your mock that return those difference and you can assert on the difference.

Does that make sense?

-------------------------------------
author: Natasha Sullivan
date: 2013-02-14 14:22:36.144000000

Patch Set 6: (4 inline comments)



Line:182, src/test/java/com/couchbase/client/CouchbaseConnectionFactoryTest.java -> no assertions here?

Line:383, src/main/java/com/couchbase/client/CouchbaseConnectionFactory.java -> please run checkstyle too when you're finished!

Line:50, src/test/java/com/couchbase/client/vbucket/ConfigurationProviderHTTPTest.java -> Good catch ,but should this be part of this changeset?

Line:7, /COMMIT_MSG -> Too long, please cut off!

-------------------------------------
