Summary: This diff ports the Preemptive Fast Fail feature to OSS. In multi threaded clients, we use a feature developed on 0.89-fb branch called Preemptive Fast Fail. This allows the client threads which would potentially fail, fail fast. The idea behind this feature is that we allow, among the hundreds of client threads, one thread to try and establish connection with the regionserver and if that succeeds, we mark it as a live node again. Meanwhile, other threads which are trying to establish connection to the same server would ideally go into the timeouts which is effectively unfruitful. We can in those cases return appropriate exceptions to those clients instead of letting them retry.
Test Plan: Unit tests
Differential Revision: https://reviews.facebook.net/D24177
Signed-off-by: stack <stack@apache.org>
Conflicts:
hbase-client/src/main/java/org/apache/hadoop/hbase/client/ClusterConnection.java
hbase-client/src/main/java/org/apache/hadoop/hbase/client/ConnectionManager.java
hbase-client/src/main/java/org/apache/hadoop/hbase/client/RpcRetryingCaller.java
Summary: This is a follow up diff for the jira to add the functionality to update configs in an online fashion.
Test Plan: Unit test. This adds an end point on the regionserver. It just tries to call the end point and make sure the call succeeds.
Differential Revision: https://reviews.facebook.net/D24861
Signed-off-by: stack <stack@apache.org>
Conflicts:
hbase-server/src/test/java/org/apache/hadoop/hbase/regionserver/TestRegionServerOnlineConfigChange.java
Summary: # Clear the cache of the server when failed
Test Plan: Add testcase `TestHTableMultiplexerFlushCache` to reproduce the bug.
Differential Revision: https://reviews.facebook.net/D24603
Signed-off-by: Elliott Clark <elliott@fb.com>
Signed-off-by: Elliott Clark <eclark@apache.org>
Fix javadoc warnings.
Fixup findbugs warnings mostly by adding annotations saying 'working as expected'.
In RpcRetryingCallerWithReadReplicas made following change which findbugs spotted:
- if (completed == null) tasks.wait();
+ while (completed == null) tasks.wait();
In RecoverableZooKeeper, made all zk accesses synchronized -- we were doing it
half-ways previously.
In RatioBasedCompactionPolicy we were making an instance of Random on
each invocation of getNextMajorCompactionTime
Conflicts:
hbase-client/src/main/java/org/apache/hadoop/hbase/client/RegionAdminServiceCallable.java
hbase-client/src/main/java/org/apache/hadoop/hbase/protobuf/ProtobufUtil.java
hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/RowTooBigException.java
Lars Hofhansl
stack
Jeffrey Zhong
Nick Dimiduk
Andrew Purtell
Enis Soztutar
Ashish Singhi
manukranthk
Matteo Bertozzi
Ted Yu
Yi Deng
Jimmy Xiang
anoopsjohn