leosunli commented on a change in pull request #2265:
URL: https://github.com/apache/hadoop/pull/2265#discussion_r483384047
##########
File path:
hadoop-hdfs-project/hadoop-hdfs-client/src/main/java/org/apache/hadoop/hdfs/DeadNodeDetector.java
##########
@@ -396,13 +395,13 @@ private void probeCallBack(Probe probe, boolean success) {
probe.getDatanodeInfo());
removeDeadNode(probe.getDatanodeInfo());
} else if (probe.getType() == ProbeType.CHECK_SUSPECT) {
- LOG.debug("Remove the node out from suspect node list: {}.",
+ LOG.info("Remove the node out from suspect node list: {}.",
Review comment:
when a lot of stale relicas, it should have many supsect nodes but
not dead nodes.
These nodes all will print this log.
What is the purpose of printing this log?
The client can access normally the suspect node.
##########
File path:
hadoop-hdfs-project/hadoop-hdfs-client/src/main/java/org/apache/hadoop/hdfs/DeadNodeDetector.java
##########
@@ -475,6 +475,7 @@ public synchronized void addNodeToDetect(DFSInputStream
dfsInputStream,
datanodeInfos.add(datanodeInfo);
}
+ LOG.warn("Add datanode {} to suspectAndDeadNodes", datanodeInfo);
Review comment:
One case: when a lot of stale relicas, will the log flood?
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
