Cassandra throws an exception when querying a very large dataset.
-----------------------------------------------------------------
Key: CASSANDRA-2707
URL: https://issues.apache.org/jira/browse/CASSANDRA-2707
Project: Cassandra
Issue Type: Bug
Affects Versions: 0.8 beta 1
Environment: Eight cassandra instances, with a replication factor of
three.
DB is running on EC2, all machines are in the same availability zone.
All machines are m1.xlarge, under 70% disk usage for the cassandra data drive,
and with 16G of RAM.
java version "1.6.0_04"
Java(TM) SE Runtime Environment (build 1.6.0_04-b12)
Java HotSpot(TM) 64-Bit Server VM (build 10.0-b19, mixed mode)
Reporter: Michael Amygdalidis
Cassandra reliably throws a runtime exception (without terminating) when
querying a very large dataset.
The cluster performs just fine in normal situations with data sets of 10,000 or
so. However, when querying a column family through either fauna/cassandra or
through CLI for all of the values matching a certain key, with a limit of 100,
the following exception is thrown.
ERROR [ReadStage:126] 2011-05-25 14:14:46,260 AbstractCassandraDaemon.java
(line 113) Fatal exception in thread Thread[ReadStage:126,5,main]
java.lang.RuntimeException: java.io.IOException: Corrupt (negative) value
length encountered
at
org.apache.cassandra.db.columniterator.IndexedSliceReader.computeNext(IndexedSliceReader.java:126)
at
org.apache.cassandra.db.columniterator.IndexedSliceReader.computeNext(IndexedSliceReader.java:49)
at
com.google.common.collect.AbstractIterator.tryToComputeNext(AbstractIterator.java:140)
at
com.google.common.collect.AbstractIterator.hasNext(AbstractIterator.java:135)
at
org.apache.cassandra.db.columniterator.SSTableSliceIterator.hasNext(SSTableSliceIterator.java:108)
at
org.apache.commons.collections.iterators.CollatingIterator.set(CollatingIterator.java:283)
at
org.apache.commons.collections.iterators.CollatingIterator.least(CollatingIterator.java:326)
at
org.apache.commons.collections.iterators.CollatingIterator.next(CollatingIterator.java:230)
at
org.apache.cassandra.utils.ReducingIterator.computeNext(ReducingIterator.java:69)
at
com.google.common.collect.AbstractIterator.tryToComputeNext(AbstractIterator.java:140)
at
com.google.common.collect.AbstractIterator.hasNext(AbstractIterator.java:135)
at
org.apache.cassandra.db.filter.SliceQueryFilter.collectReducedColumns(SliceQueryFilter.java:116)
at
org.apache.cassandra.db.filter.QueryFilter.collectCollatedColumns(QueryFilter.java:130)
at
org.apache.cassandra.db.ColumnFamilyStore.getTopLevelColumns(ColumnFamilyStore.java:1302)
at
org.apache.cassandra.db.ColumnFamilyStore.getColumnFamily(ColumnFamilyStore.java:1187)
at
org.apache.cassandra.db.ColumnFamilyStore.getColumnFamily(ColumnFamilyStore.java:1144)
at org.apache.cassandra.db.Table.getRow(Table.java:385)
at
org.apache.cassandra.db.SliceFromReadCommand.getRow(SliceFromReadCommand.java:61)
at
org.apache.cassandra.db.ReadVerbHandler.doVerb(ReadVerbHandler.java:69)
at
org.apache.cassandra.net.MessageDeliveryTask.run(MessageDeliveryTask.java:72)
at
java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:885)
at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:907)
at java.lang.Thread.run(Thread.java:619)
Caused by: java.io.IOException: Corrupt (negative) value length encountered
at
org.apache.cassandra.utils.ByteBufferUtil.readWithLength(ByteBufferUtil.java:348)
at
org.apache.cassandra.db.ColumnSerializer.deserialize(ColumnSerializer.java:126)
at
org.apache.cassandra.db.ColumnSerializer.deserialize(ColumnSerializer.java:82)
at
org.apache.cassandra.db.ColumnSerializer.deserialize(ColumnSerializer.java:72)
at
org.apache.cassandra.db.ColumnSerializer.deserialize(ColumnSerializer.java:36)
at
org.apache.cassandra.db.columniterator.IndexedSliceReader$IndexedBlockFetcher.getNextBlock(IndexedSliceReader.java:179)
at
org.apache.cassandra.db.columniterator.IndexedSliceReader.computeNext(IndexedSliceReader.java:121)
... 22 more
Additional Info:
* I can confirm that the same exception is reliably thrown on three instances
at about the same as the query is executed.
* The timeout for a remote procedure call is between nodes is 10 seconds, which
is about the time it takes for the query to respond with null.
* Asking for forward or reverse search does not affect results, however, in
production we'd need to do a reverse search.
Steps to Reproduce:
Have a column family with at least 100 million values, including at least 30
million with the same key. Try to get 100 items of a given key from that
column family.
Expected behaviour: To get back the 100 items we queried for, which is what
happens when the number of items under a given key is not so large. The
unexpected behaviour only manifests itself when the number of possible items is
extremely large.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
