Varun Thacker created SOLR-8335:
-----------------------------------
Summary: HdfsLockFactory does not allow core to come up after a
node was killed
Key: SOLR-8335
URL: https://issues.apache.org/jira/browse/SOLR-8335
Project: Solr
Issue Type: Bug
Affects Versions: 5.3, 5.2.1, 5.2, 5.1, 5.0, 5.3.1
Reporter: Varun Thacker
When using HdfsLockFactory if a node gets killed instead of a graceful shutdown
the write.lock file remains in HDFS . The next time you start the node the core
doesn't load up because of LockObtainFailedException .
I was able to reproduce this in all 5.x versions of Solr . The problem wasn't
there when I tested it in 4.10.4
Steps to reproduce this on 5.x
1. Create directory in HDFS : {{bin/hdfs dfs -mkdir /solr}}
2. Start Solr: {{bin/solr start -Dsolr.directoryFactory=HdfsDirectoryFactory
-Dsolr.lock.type=hdfs -Dsolr.data.dir=hdfs://localhost:9000/solr
-Dsolr.updatelog=hdfs://localhost:9000/solr}}
3. Create core: {{./bin/solr create -c test -n data_driven}}
4. Kill solr
5. The lock file is there in HDFS and is called {{write.lock}}
6. Start Solr again and you get a stack trace like this:
{code}
2015-11-23 13:28:04.287 ERROR (coreLoadExecutor-6-thread-1) [ x:test]
o.a.s.c.CoreContainer Error creating core [test]: Index locked for write for
core 'test'. Solr now longer supports forceful unlocking via 'unlockOnStartup'.
Please verify locks manually!
org.apache.solr.common.SolrException: Index locked for write for core 'test'.
Solr now longer supports forceful unlocking via 'unlockOnStartup'. Please
verify locks manually!
at org.apache.solr.core.SolrCore.<init>(SolrCore.java:820)
at org.apache.solr.core.SolrCore.<init>(SolrCore.java:659)
at org.apache.solr.core.CoreContainer.create(CoreContainer.java:723)
at org.apache.solr.core.CoreContainer$1.call(CoreContainer.java:443)
at org.apache.solr.core.CoreContainer$1.call(CoreContainer.java:434)
at java.util.concurrent.FutureTask.run(FutureTask.java:266)
at
org.apache.solr.common.util.ExecutorUtil$MDCAwareThreadPoolExecutor$1.run(ExecutorUtil.java:210)
at
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
at java.lang.Thread.run(Thread.java:745)
Caused by: org.apache.lucene.store.LockObtainFailedException: Index locked for
write for core 'test'. Solr now longer supports forceful unlocking via
'unlockOnStartup'. Please verify locks manually!
at org.apache.solr.core.SolrCore.initIndex(SolrCore.java:528)
at org.apache.solr.core.SolrCore.<init>(SolrCore.java:761)
... 9 more
2015-11-23 13:28:04.289 ERROR (coreContainerWorkExecutor-2-thread-1) [ ]
o.a.s.c.CoreContainer Error waiting for SolrCore to be created
java.util.concurrent.ExecutionException: org.apache.solr.common.SolrException:
Unable to create core [test]
at java.util.concurrent.FutureTask.report(FutureTask.java:122)
at java.util.concurrent.FutureTask.get(FutureTask.java:192)
at org.apache.solr.core.CoreContainer$2.run(CoreContainer.java:472)
at
java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
at java.util.concurrent.FutureTask.run(FutureTask.java:266)
at
org.apache.solr.common.util.ExecutorUtil$MDCAwareThreadPoolExecutor$1.run(ExecutorUtil.java:210)
at
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
at java.lang.Thread.run(Thread.java:745)
Caused by: org.apache.solr.common.SolrException: Unable to create core [test]
at org.apache.solr.core.CoreContainer.create(CoreContainer.java:737)
at org.apache.solr.core.CoreContainer$1.call(CoreContainer.java:443)
at org.apache.solr.core.CoreContainer$1.call(CoreContainer.java:434)
... 5 more
Caused by: org.apache.solr.common.SolrException: Index locked for write for
core 'test'. Solr now longer supports forceful unlocking via 'unlockOnStartup'.
Please verify locks manually!
at org.apache.solr.core.SolrCore.<init>(SolrCore.java:820)
at org.apache.solr.core.SolrCore.<init>(SolrCore.java:659)
at org.apache.solr.core.CoreContainer.create(CoreContainer.java:723)
... 7 more
Caused by: org.apache.lucene.store.LockObtainFailedException: Index locked for
write for core 'test'. Solr now longer supports forceful unlocking via
'unlockOnStartup'. Please verify locks manually!
at org.apache.solr.core.SolrCore.initIndex(SolrCore.java:528)
at org.apache.solr.core.SolrCore.<init>(SolrCore.java:761)
... 9 more
{code}
In 4.10.4 I saw these two differences
1. The lock file name was different . It's something like :
{{/solr/index/HdfsDirectory@46ad6bd3
lockFactory=org.apache.solr.store.hdfs.hdfslockfact...@4b44b5f6-write.lock}}
2. When the node is started again after it was killed , it loaded up the core
just fine but there were two lock files in hdfs now . 4b44b5f6-write.lock is
the latest one
{code}
/solr/index/HdfsDirectory@46ad6bd3
lockFactory=org.apache.solr.store.hdfs.hdfslockfact...@4b44b5f6-write.lock
/solr/index/HdfsDirectory@52959724
lockFactory=org.apache.solr.store.hdfs.hdfslockfact...@9d59d3f-write.lock
{code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
