Doug Rohrer
7e7cb73d55
Add rt:wait_until_bucket_type_visible to avoid floppiness in test - verify would fail because, while handoff had completed, the bucket type information hadn't yet been gossiped to the new node.
2015-07-28 11:03:45 -04:00
Fred Dushin
5cf5ef0d7a
Cleaned up commented code. No change in behavior.
2015-07-24 12:12:31 -04:00
Fred Dushin
87f3f88d49
Modified test to wait to send messages until all waiting vnodes are in the handoff state. Also corrected some of the arithmetic errors in counting the number of entries we were supposed to send. (doh)
2015-07-24 11:07:15 -04:00
Fred Dushin
6f979d69ec
Mode a change in the protocol between vnodes and our bg proc in the test to keep vnodes in the handoff state until all writes have completed.
...
Plus some miscellaneous tightening of test conditions.
2015-07-23 17:47:47 -04:00
Fred Dushin
c8d3138cfc
Missed a change in previous commit
2015-07-23 14:22:12 -04:00
Fred Dushin
399759c330
Added some documentation to the test.
...
Also changed the name of the internal identifier for the globally registered bg process, to avoid potential name collisions in the future.
2015-07-23 14:18:25 -04:00
Fred Dushin
a418bae347
Write-Once test improvements
...
* Made tests of handoff conditions more strict than they were before
* Added an intercept to the async write_once test to test that asynchronous writes actually happened
* Added a test case for the riak_kv (advanced) allow_async_put configuration variable.
2015-07-23 12:25:35 -04:00
Fred Dushin
666f0ca375
Added intercepts to count the synchronous and asynchronous calls through handle_command, to ensure we are testing the case statements in riak_kv_vnode:handle_command with write_once puts.
2015-07-20 21:23:22 -04:00
Fred Dushin
e38c0f2533
Cleaned up async process to remove code that is not exercised by this test.
2015-07-17 09:04:17 -04:00
Fred Dushin
85c79f48af
Cleaned up the verify_handoff_write_once to run faster
...
(by not waiting for handoffs, just pending changes, as the old test did)
2015-07-16 17:21:10 -04:00
Fred Dushin
c2d610a6d8
Force calls to riak_kv_vnode:handle_handoff_command during a handoff.
2015-07-16 17:03:37 -04:00
Fred Dushin
3fa57b4025
Added a cap to the number of messages to be send asynchronously, to give handoff a chance to complete.
2015-07-07 17:25:49 -04:00
Fred Dushin
fbedd4cd33
Tried to make the verify_handoff_write_once test more predictable
2015-07-07 16:45:05 -04:00
Fred Dushin
f330f867b2
Reverted to original verify_handoff_write_once test with modifications
...
* Backed out the change to riak_core_handoff_sender_intercepts which had
a modified intercept
* Modified the handle_command intercept to ensure we don't count the
vnode crashes that were happening before the fix
* Reverted the test to the original, but check to make sure we can read
all the entries, including the ones we added during handoff
2015-07-02 11:48:20 -04:00
Fred Dushin
06836392cd
Re-write of handle_handoff_write_once to test handle_handoff_command.
...
Still needs some work!
2015-07-02 10:59:01 -04:00
Fred Dushin
d34ff09f8e
Updated write_once handoff test to use riak_kv_w1c_put_req_v1 record, and relaxed check in test to avoid potential race condition.
2015-04-01 16:56:36 -04:00
Fred Dushin
5f547a8364
Name changes. Test is still broken though, due to use of #riak_kv_w1c_put_req_v1 record.
2015-03-27 20:57:09 -04:00
Fred Dushin
57638a86a3
Name change fast_path -> write_once
2015-03-27 20:38:47 -04:00
