Ivan Bessonov created IGNITE-11683:
--------------------------------------
Summary: DistributedMetaStoragePersistentTest#testClientReconnect
hangs sometimes.
Key: IGNITE-11683
URL: https://issues.apache.org/jira/browse/IGNITE-11683
Project: Ignite
Issue Type: Test
Reporter: Ivan Bessonov
Assignee: Ivan Bessonov
The problem occurs right after this line:
{code:java}
assertTrue(GridTestUtils.waitForCondition(() ->
metastorage(1).getUpdatesCount() == expUpdatesCnt, 15_000));
{code}
Client node might not be fully reconnected yet. Adding following line resolves
the problem in the particular test:
{code:java}
grid(1).cluster().clientReconnectFuture().get();
{code}
I don't consider this a proper fix. Stopping the client that hasn't finished
its reconnect shouldn't result in infinite waiting (or deadlock). Client node
should be stopped successfully.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
