[
https://issues.apache.org/jira/browse/HDFS-17030?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17752280#comment-17752280
]
ASF GitHub Bot commented on HDFS-17030:
---------------------------------------
xinglin commented on code in PR #5878:
URL: https://github.com/apache/hadoop/pull/5878#discussion_r1288009483
##########
hadoop-hdfs-project/hadoop-hdfs-client/src/main/java/org/apache/hadoop/hdfs/server/namenode/ha/ObserverReadProxyProvider.java:
##########
@@ -285,13 +323,67 @@ private synchronized NNProxyInfo<T>
changeProxy(NNProxyInfo<T> initial) {
}
currentIndex = (currentIndex + 1) % nameNodeProxies.size();
currentProxy = createProxyIfNeeded(nameNodeProxies.get(currentIndex));
- currentProxy.setCachedState(getHAServiceState(currentProxy));
+ currentProxy.setCachedState(getHAServiceStateWithTimeout(currentProxy));
LOG.debug("Changed current proxy from {} to {}",
initial == null ? "none" : initial.proxyInfo,
currentProxy.proxyInfo);
return currentProxy;
}
+ /**
+ * Execute getHAServiceState() call with a timeout, to avoid a long wait when
+ * an NN becomes irresponsive to rpc requests
+ * (when a thread/heap dump is being taken, e.g.).
+ *
+ * For each getHAServiceState() call, a task is created and submitted to a
+ * threadpool for execution. We will wait for a response up to
+ * namenodeHAStateProbeTimeoutSec and cancel these requests if they time out.
+ *
+ * The implementation is split into two functions so that we can unit test
+ * the second function.
+ */
+ HAServiceState getHAServiceStateWithTimeout(final NNProxyInfo<T> proxyInfo) {
+ Callable<HAServiceState> getHAServiceStateTask = () ->
getHAServiceState(proxyInfo);
+
+ try {
+ Future<HAServiceState> task =
+ nnProbingThreadPool.submit(getHAServiceStateTask);
+ return getHAServiceStateWithTimeout(proxyInfo, task);
+ } catch (RejectedExecutionException e) {
+ LOG.warn("Run out of threads to submit the request to query HA state. "
+ + "Ok to return null and we will fallback to use active NN to serve "
+ + "this request.");
+ return null;
+ }
+ }
+
+ HAServiceState getHAServiceStateWithTimeout(final NNProxyInfo<T> proxyInfo,
+ Future<HAServiceState> task) {
+ HAServiceState state = null;
+ try {
+ if (namenodeHAStateProbeTimeoutMs > 0) {
+ state = task.get(namenodeHAStateProbeTimeoutMs, TimeUnit.MILLISECONDS);
+ } else {
+ // Disable timeout by waiting indefinitely when
namenodeHAStateProbeTimeoutSec is set to 0
+ // or a negative value.
+ state = task.get();
+ }
+ LOG.debug("HA State for {} is {}", proxyInfo.proxyInfo, state);
+ } catch (TimeoutException e) {
+ // Cancel the task on timeout
+ String msg = String.format("Cancel NN probe task due to timeout for %s",
proxyInfo.proxyInfo);
+ LOG.warn(msg, e);
+ if (task != null) {
Review Comment:
removed.
> Limit wait time for getHAServiceState in ObserverReaderProxy
> ------------------------------------------------------------
>
> Key: HDFS-17030
> URL: https://issues.apache.org/jira/browse/HDFS-17030
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: hdfs
> Affects Versions: 3.4.0
> Reporter: Xing Lin
> Assignee: Xing Lin
> Priority: Minor
> Labels: pull-request-available
> Fix For: 3.4.0
>
>
> When namenode HA is enabled and a standby NN is not responsible, we have
> observed it would take a long time to serve a request, even though we have a
> healthy observer or active NN.
> Basically, when a standby is down, the RPC client would (re)try to create
> socket connection to that standby for _ipc.client.connect.timeout_ _*
> ipc.client.connect.max.retries.on.timeouts_ before giving up. When we take a
> heap dump at a standby, the NN still accepts the socket connection but it
> won't send responses to these RPC requests and we would timeout after
> _ipc.client.rpc-timeout.ms._ This adds a significantly latency. For clusters
> at Linkedin, we set _ipc.client.rpc-timeout.ms_ to 120 seconds and thus a
> request takes more than 2 mins to complete when we take a heap dump at a
> standby. This has been causing user job failures.
> We could set _ipc.client.rpc-timeout.ms to_ a smaller value when sending
> getHAServiceState requests in ObserverReaderProxy (for user rpc requests, we
> still use the original value from the config). However, that would double the
> socket connection between clients and the NN (which is a deal-breaker).
> The proposal is to add a timeout on getHAServiceState() calls in
> ObserverReaderProxy and we will only wait for the timeout for an NN to
> respond its HA state. Once we pass that timeout, we will move on to probe the
> next NN.
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
