[
https://issues.apache.org/jira/browse/CASSANDRA-15526?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17061600#comment-17061600
]
ZhaoYang commented on CASSANDRA-15526:
--------------------------------------
{quote}The issue is that in v11 there's a potential race condition in which the
head node may have been initialized, but the LongAdder hasn't been incremented
yet, which leaves it briefly in an inconsistent state.
{quote}
{quote}I confirmed that's the case by logging the value of keys.size() in
TrieMemIndex#search right before and after the keys.isEmpty() check. The size
is 0 at runtime, but when the IDE debugger inspects the value, it has changed
to 1 already.
{quote}
Good catch! Should we fix {{SkipListMemIndex}} as well? we can also pass the
size via {{KeyRangeIterator}} constructor, so that it's computed once.
In jdk 11 {{ConcurrentSkipListMap#size()}} javadoc:
{quote} * <p>Beware that, unlike in most collections, this method is
* <em>NOT</em> a constant-time operation. Because of the
* asynchronous nature of these sets, determining the current
* number of elements requires traversing them all to count them.
* Additionally, it is possible for the size to change during
* execution of this method, in which case the returned result
* will be inaccurate. Thus, this method is typically not very
* useful in concurrent applications.{quote}
I think it's safe not to include keys when size()=0 but isEmpty()=false, as
write is not considered successful from client POV.
> Fix flakey test - org.apache.cassandra.index.sasi.SASIIndexTest
> testConcurrentMemtableReadsAndWrites
> ----------------------------------------------------------------------------------------------------
>
> Key: CASSANDRA-15526
> URL: https://issues.apache.org/jira/browse/CASSANDRA-15526
> Project: Cassandra
> Issue Type: Bug
> Components: Test/unit
> Reporter: David Capwell
> Assignee: Gianluca Righetto
> Priority: Normal
> Labels: pull-request-available
> Fix For: 4.0-alpha
>
> Attachments: 15526-trunk-4.0.txt, jvm_15526.zip, unit_tests_15526
>
> Time Spent: 10m
> Remaining Estimate: 0h
>
> {code}
> junit.framework.AssertionFailedError
> at
> org.apache.cassandra.index.sasi.utils.RangeIterator.<init>(RangeIterator.java:46)
> at
> org.apache.cassandra.index.sasi.memory.KeyRangeIterator.<init>(KeyRangeIterator.java:42)
> at
> org.apache.cassandra.index.sasi.memory.TrieMemIndex$ConcurrentTrie.search(TrieMemIndex.java:150)
> at
> org.apache.cassandra.index.sasi.memory.TrieMemIndex.search(TrieMemIndex.java:102)
> at
> org.apache.cassandra.index.sasi.memory.IndexMemtable.search(IndexMemtable.java:70)
> at
> org.apache.cassandra.index.sasi.conf.ColumnIndex.searchMemtable(ColumnIndex.java:138)
> at org.apache.cassandra.index.sasi.TermIterator.build(TermIterator.java:91)
> at
> org.apache.cassandra.index.sasi.plan.QueryController.getIndexes(QueryController.java:145)
> at
> org.apache.cassandra.index.sasi.plan.Operation$Builder.complete(Operation.java:434)
> at org.apache.cassandra.index.sasi.plan.QueryPlan.analyze(QueryPlan.java:57)
> at org.apache.cassandra.index.sasi.plan.QueryPlan.execute(QueryPlan.java:68)
> at
> org.apache.cassandra.index.sasi.SASIIndex.lambda$searcherFor$2(SASIIndex.java:301)
> at org.apache.cassandra.db.ReadCommand.executeLocally(ReadCommand.java:455)
> at
> org.apache.cassandra.index.sasi.SASIIndexTest.getIndexed(SASIIndexTest.java:2576)
> at
> org.apache.cassandra.index.sasi.SASIIndexTest.getPaged(SASIIndexTest.java:2537)
> at
> org.apache.cassandra.index.sasi.SASIIndexTest.testConcurrentMemtableReadsAndWrites(SASIIndexTest.java:1108)
> at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native
> Method)
> at
> java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
> at
> java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
> {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]
