[ 
https://issues.apache.org/jira/browse/CASSANDRA-17145?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17445100#comment-17445100
 ] 

Berenguer Blasi edited comment on CASSANDRA-17145 at 11/17/21, 10:57 AM:
-------------------------------------------------------------------------

[~samt] I need to pick your brain here for a test failure introduced by CEP-10 
which should be quick for you to asses (I hope) bc that PR is massive.

As you can see 
[here|https://app.circleci.com/pipelines/github/bereng/cassandra/502/workflows/f5880a6d-4742-4607-9294-cac78dd36f11/jobs/4561/steps]
 these tests started failing in trunk. Drilling a bit I found the culprit being 
[this 
line|https://github.com/apache/cassandra/blob/cassandra-4.0/src/java/org/apache/cassandra/db/ColumnFamilyStore.java#L2877]
 in 4.0 using {{numberOfExecutors}} in the loop whereas trunk 
[here|https://github.com/apache/cassandra/blame/trunk/src/java/org/apache/cassandra/db/ColumnFamilyStore.java#L2944]
 using {{DatabaseDescriptor.getAllDataFileLocations().length}}. This produces 
the following error in 500+ tests:

{noformat}
[junit-timeout] java.lang.ExceptionInInitializerError
[junit-timeout]         at 
org.apache.cassandra.db.Keyspace.initCf(Keyspace.java:472)
[junit-timeout]         at 
org.apache.cassandra.db.Keyspace.<init>(Keyspace.java:366)
[junit-timeout]         at 
org.apache.cassandra.db.Keyspace.open(Keyspace.java:154)
[junit-timeout]         at 
org.apache.cassandra.db.Keyspace.open(Keyspace.java:130)
[junit-timeout]         at 
org.apache.cassandra.db.Keyspace.openAndGetStore(Keyspace.java:189)
[junit-timeout]         at 
org.apache.cassandra.index.IndexRegistry.obtain(IndexRegistry.java:251)
[junit-timeout]         at 
org.apache.cassandra.db.partitions.PartitionUpdate.validateIndexedColumns(PartitionUpdate.java:473)
[junit-timeout]         at 
org.apache.cassandra.db.IMutation.validateIndexedColumns(IMutation.java:42)
[junit-timeout]         at 
org.apache.cassandra.cql3.statements.SingleTableUpdatesCollector.toMutations(SingleTableUpdatesCollector.java:107)
[junit-timeout]         at 
org.apache.cassandra.cql3.statements.ModificationStatement.getMutations(ModificationStatement.java:690)
[junit-timeout]         at 
org.apache.cassandra.cql3.statements.ModificationStatement.executeInternalWithoutCondition(ModificationStatement.java:633)
[junit-timeout]         at 
org.apache.cassandra.cql3.statements.ModificationStatement.executeLocally(ModificationStatement.java:625)
[junit-timeout]         at 
org.apache.cassandra.cql3.QueryProcessor.executeOnceInternal(QueryProcessor.java:453)
[junit-timeout]         at 
org.apache.cassandra.cql3.QueryProcessor.executeOnceInternal(QueryProcessor.java:435)
[junit-timeout]         at 
org.apache.cassandra.db.SystemKeyspace.persistLocalMetadata(SystemKeyspace.java:491)
[junit-timeout]         at 
org.apache.cassandra.ServerTestUtils.prepareServer(ServerTestUtils.java:129)
[junit-timeout]         at 
org.apache.cassandra.cql3.CQLTester.prepareServer(CQLTester.java:263)
[junit-timeout]         at 
org.apache.cassandra.cql3.CQLTester.setUpClass(CQLTester.java:311)
[junit-timeout] Caused by: java.lang.ArrayIndexOutOfBoundsException: 1
[junit-timeout]         at 
org.apache.cassandra.db.ColumnFamilyStore$PerDiskFlushExecutors.createPerDiskFlushWriters(ColumnFamilyStore.java:2946)
[junit-timeout]         at 
org.apache.cassandra.db.ColumnFamilyStore$PerDiskFlushExecutors.<init>(ColumnFamilyStore.java:2934)
[junit-timeout]         at 
org.apache.cassandra.db.ColumnFamilyStore.<clinit>(ColumnFamilyStore.java:148)
{noformat}

Changing it back to {{numberOfExecutors}} seems to fix the problem and seems 
correct at face value without knowing anything on CEP-10 implementation 
details. Can you confirm the check is good please?


was (Author: bereng):
[~samt] I need to pick your brain here for a test failure introduced by CEP-10 
which should be quick for you to asses (I hope) bc that PR is massive.

As you can see 
[here|https://app.circleci.com/pipelines/github/bereng/cassandra/502/workflows/f5880a6d-4742-4607-9294-cac78dd36f11/jobs/4561/steps]
 these tests started failing. Drilling a bit I found the culprit being [this 
line|https://github.com/apache/cassandra/blob/cassandra-4.0/src/java/org/apache/cassandra/db/ColumnFamilyStore.java#L2877]
 in 4.0 using {{numberOfExecutors}} in the loop whereas trunk 
[here|https://github.com/apache/cassandra/blame/trunk/src/java/org/apache/cassandra/db/ColumnFamilyStore.java#L2944]
 using {{DatabaseDescriptor.getAllDataFileLocations().length}}. This produces 
the following error in 500+ tests:

{noformat}
[junit-timeout] java.lang.ExceptionInInitializerError
[junit-timeout]         at 
org.apache.cassandra.db.Keyspace.initCf(Keyspace.java:472)
[junit-timeout]         at 
org.apache.cassandra.db.Keyspace.<init>(Keyspace.java:366)
[junit-timeout]         at 
org.apache.cassandra.db.Keyspace.open(Keyspace.java:154)
[junit-timeout]         at 
org.apache.cassandra.db.Keyspace.open(Keyspace.java:130)
[junit-timeout]         at 
org.apache.cassandra.db.Keyspace.openAndGetStore(Keyspace.java:189)
[junit-timeout]         at 
org.apache.cassandra.index.IndexRegistry.obtain(IndexRegistry.java:251)
[junit-timeout]         at 
org.apache.cassandra.db.partitions.PartitionUpdate.validateIndexedColumns(PartitionUpdate.java:473)
[junit-timeout]         at 
org.apache.cassandra.db.IMutation.validateIndexedColumns(IMutation.java:42)
[junit-timeout]         at 
org.apache.cassandra.cql3.statements.SingleTableUpdatesCollector.toMutations(SingleTableUpdatesCollector.java:107)
[junit-timeout]         at 
org.apache.cassandra.cql3.statements.ModificationStatement.getMutations(ModificationStatement.java:690)
[junit-timeout]         at 
org.apache.cassandra.cql3.statements.ModificationStatement.executeInternalWithoutCondition(ModificationStatement.java:633)
[junit-timeout]         at 
org.apache.cassandra.cql3.statements.ModificationStatement.executeLocally(ModificationStatement.java:625)
[junit-timeout]         at 
org.apache.cassandra.cql3.QueryProcessor.executeOnceInternal(QueryProcessor.java:453)
[junit-timeout]         at 
org.apache.cassandra.cql3.QueryProcessor.executeOnceInternal(QueryProcessor.java:435)
[junit-timeout]         at 
org.apache.cassandra.db.SystemKeyspace.persistLocalMetadata(SystemKeyspace.java:491)
[junit-timeout]         at 
org.apache.cassandra.ServerTestUtils.prepareServer(ServerTestUtils.java:129)
[junit-timeout]         at 
org.apache.cassandra.cql3.CQLTester.prepareServer(CQLTester.java:263)
[junit-timeout]         at 
org.apache.cassandra.cql3.CQLTester.setUpClass(CQLTester.java:311)
[junit-timeout] Caused by: java.lang.ArrayIndexOutOfBoundsException: 1
[junit-timeout]         at 
org.apache.cassandra.db.ColumnFamilyStore$PerDiskFlushExecutors.createPerDiskFlushWriters(ColumnFamilyStore.java:2946)
[junit-timeout]         at 
org.apache.cassandra.db.ColumnFamilyStore$PerDiskFlushExecutors.<init>(ColumnFamilyStore.java:2934)
[junit-timeout]         at 
org.apache.cassandra.db.ColumnFamilyStore.<clinit>(ColumnFamilyStore.java:148)
{noformat}

Changing that seems to fix the problem and seems correct at face value without 
knowing anything on CEP-10 implementation details. Can you confirm the check is 
good please?

> utests_system_keyspace_directory failing and not run in Jenkins
> ---------------------------------------------------------------
>
>                 Key: CASSANDRA-17145
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-17145
>             Project: Cassandra
>          Issue Type: Improvement
>          Components: CI
>            Reporter: Ekaterina Dimitrova
>            Assignee: Berenguer Blasi
>            Priority: Normal
>             Fix For: 4.1
>
>
> It seems to me utests_system_keyspace_directory which we run with 4.0 and 
> trunk in CircleCI do not run in Jenkins as they all fail in trunk now in 
> CircleCI and nothing in Jenkins. The issue is not environmental and I don't 
> see that group of tests under Stage View in Jenkins so they are probably not 
> added there? We need to investigate that further.
> CC [~mck]  and [~blerer] if they know anything about those - like work in 
> progress to be added or a reason not to be added.
> FYI [~bereng] 



--
This message was sent by Atlassian Jira
(v8.20.1#820001)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to