[
https://issues.apache.org/jira/browse/HBASE-26027?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17385922#comment-17385922
]
Michael Stack commented on HBASE-26027:
---------------------------------------
Let me follow [~apurtell] 's lead. I tried the test on 2.3.6 and it fails for
me on mac osx too:
{code:java}
[INFO] Running org.apache.hadoop.hbase.TestClientOperationTimeout
[ERROR] Tests run: 4, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 18.488
s <<< FAILURE! - in org.apache.hadoop.hbase.TestClientOperationTimeout
[ERROR] org.apache.hadoop.hbase.TestClientOperationTimeout.testMultiPutsTimeout
Time elapsed: 0.629 s <<< FAILURE!
java.lang.AssertionError: should not reach here
at
org.apache.hadoop.hbase.TestClientOperationTimeout.testMultiPutsTimeout(TestClientOperationTimeout.java:176)[INFO]
[INFO] Results:
[INFO]
[ERROR] Failures:
[ERROR] TestClientOperationTimeout.testMultiPutsTimeout:176 should not reach
here
[INFO]
[ERROR] Tests run: 4, Failures: 1, Errors: 0, Skipped: 0 {code}
Let me revert from branch-2.3 too... and might as well do 2.5 too while we are
at it.
> The calling of HTable.batch blocked at AsyncRequestFutureImpl.waitUntilDone
> caused by ArrayStoreException
> ---------------------------------------------------------------------------------------------------------
>
> Key: HBASE-26027
> URL: https://issues.apache.org/jira/browse/HBASE-26027
> Project: HBase
> Issue Type: Bug
> Components: Client
> Affects Versions: 2.2.7, 2.3.5, 2.4.4
> Reporter: Zheng Wang
> Assignee: Zheng Wang
> Priority: Major
> Fix For: 2.5.0, 2.3.6, 2.4.6
>
>
> The batch api of HTable contains a param named results to store result or
> exception, its type is Object[].
> If user pass an array with other type, eg:
> org.apache.hadoop.hbase.client.Result, and if we need to put an exception
> into it by some reason, then the ArrayStoreException will occur in
> AsyncRequestFutureImpl.updateResult, then the
> AsyncRequestFutureImpl.decActionCounter will be skipped, then in the
> AsyncRequestFutureImpl.waitUntilDone we will stuck at here checking the
> actionsInProgress again and again, forever.
> It is better to add an cutoff calculated by operationTimeout, instead of only
> depend on the value of actionsInProgress.
> BTW, this issue only for 2.x, since 3.x the implement has refactored.
> How to reproduce:
> 1: add sleep in RSRpcServices.multi to mock slow response
> {code:java}
> try {
> Thread.sleep(2000);
> } catch (InterruptedException e) {
> e.printStackTrace();
> }
> {code}
> 2: set time out in config
> {code:java}
> conf.set("hbase.rpc.timeout","2000");
> conf.set("hbase.client.operation.timeout","6000");
> {code}
> 3: call batch api
> {code:java}
> Table table = HbaseUtil.getTable("test");
> byte[] cf = Bytes.toBytes("f");
> byte[] c = Bytes.toBytes("c1");
> List<Get> gets = new ArrayList<>();
> for (int i = 0; i < 10; i++) {
> byte[] rk = Bytes.toBytes("rk-" + i);
> Get get = new Get(rk);
> get.addColumn(cf, c);
> gets.add(get);
> }
> Result[] results = new Result[gets.size()];
> table.batch(gets, results);
> {code}
> The log will looks like below:
> {code:java}
> [ERROR] [2021/06/22 23:23:00,676] hconnection-0x6b927fb-shared-pool3-t1 -
> id=1 error for test processing localhost,16020,1624343786295
> java.lang.ArrayStoreException: org.apache.hadoop.hbase.DoNotRetryIOException
> at
> org.apache.hadoop.hbase.client.AsyncRequestFutureImpl.updateResult(AsyncRequestFutureImpl.java:1242)
> at
> org.apache.hadoop.hbase.client.AsyncRequestFutureImpl.trySetResultSimple(AsyncRequestFutureImpl.java:1087)
> at
> org.apache.hadoop.hbase.client.AsyncRequestFutureImpl.setError(AsyncRequestFutureImpl.java:1021)
> at
> org.apache.hadoop.hbase.client.AsyncRequestFutureImpl.manageError(AsyncRequestFutureImpl.java:683)
> at
> org.apache.hadoop.hbase.client.AsyncRequestFutureImpl.receiveGlobalFailure(AsyncRequestFutureImpl.java:716)
> at
> org.apache.hadoop.hbase.client.AsyncRequestFutureImpl.access$1500(AsyncRequestFutureImpl.java:69)
> at
> org.apache.hadoop.hbase.client.AsyncRequestFutureImpl$SingleServerRequestRunnable.run(AsyncRequestFutureImpl.java:219)
> at
> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
> at java.util.concurrent.FutureTask.run$$$capture(FutureTask.java:266)
> at java.util.concurrent.FutureTask.run(FutureTask.java)
> at
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
> at
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
> at java.lang.Thread.run(Thread.java:748)
> [INFO ] [2021/06/22 23:23:10,375] main - #1, waiting for 10 actions to
> finish on table: test
> [INFO ] [2021/06/22 23:23:20,378] main - #1, waiting for 10 actions to
> finish on table: test
> [INFO ] [2021/06/22 23:23:30,384] main - #1, waiting for 10 actions to
> finish on table:
> [INFO ] [2021/06/22 23:23:40,387] main - #1, waiting for 10 actions to
> finish on table: test
> [INFO ] [2021/06/22 23:23:50,397] main - #1, waiting for 10 actions to
> finish on table: test
> [INFO ] [2021/06/22 23:24:00,400] main - #1, waiting for 10 actions to
> finish on table: test
> [INFO ] [2021/06/22 23:24:10,408] main - #1, waiting for 10 actions to
> finish on table: test
> [INFO ] [2021/06/22 23:24:20,413] main - #1, waiting for 10 actions to
> finish on table: test
> {code}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
