[
https://issues.apache.org/jira/browse/HDFS-15019?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16983619#comment-16983619
]
Yiqun Lin commented on HDFS-15019:
----------------------------------
We can put common setting in @Before method and leave specific setting in test
method. Here io.bytes.per.checksum is a deprecated key, use
{{HdfsClientConfigKeys.DFS_BYTES_PER_CHECKSUM_KEY}} instead of.
{code}
@Before
public void setUp() {
cluster = null;
conf = new HdfsConfiguration();
conf.setBoolean(DFS_CLIENT_DEAD_NODE_DETECTION_ENABLED_KEY, true);
conf.setLong(DFS_CLIENT_DEAD_NODE_DETECTION_PROBE_DEAD_NODE_INTERVAL_MS_KEY,
1000);
conf.setLong(
DFS_CLIENT_DEAD_NODE_DETECTION_PROBE_SUSPECT_NODE_INTERVAL_MS_KEY, 100);
// We'll be using a 512 bytes block size just for tests
// so making sure the checksum bytes match it too.
conf.setInt(HdfsClientConfigKeys.DFS_BYTES_PER_CHECKSUM_KEY, 512);
}
{code}
> Refactor the unit test of TestDeadNodeDetection
> ------------------------------------------------
>
> Key: HDFS-15019
> URL: https://issues.apache.org/jira/browse/HDFS-15019
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Reporter: Yiqun Lin
> Assignee: Lisheng Sun
> Priority: Minor
>
> There are many duplicated lines in unit test \{{TestDeadNodeDetection}}. We
> can simplified that.
> In additional in {{testDeadNodeDetectionInMultipleDFSInputStream}}, the
> DFSInputstream is passed incorrectly in asset operation.
> {code}
> din2 = (DFSInputStream) in1.getWrappedStream();
> {code}
> Should be
> {code}
> din2 = (DFSInputStream) in2.getWrappedStream();
> {code}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
