[CASSANDRA-6531] Failure to start after unclean shutdown - java.lang.IllegalArgumentException: bufferSize must be positive

• Type: Bug
• Status: Resolved
• Priority: Normal
• Resolution: Fixed
• Fix Version/s: 1.2.14, 2.0.5
• Labels: compression

Environment

Cassandra 2.0.3 with patches for CASSANDRA-6496 and CASSANDRA-6284, built from source; Linux; XFS on the data disks; 5 data disks; 6 nodes

• Severity: Normal
• Since Version: 1.0.0
• Assignee: Jonathan Ellis
• Reporter: Nikolai Grigoriev

Authors

Jonathan Ellis

Reviewers

Pavel Yaskevich

• Created: 27/Dec/13 20:53
• Updated: 16/Apr/19 09:31
• Resolved: 28/Dec/13 06:18

Description

We had a severe power outage in the lab that resulted in unclean shutdown of the Cassandra servers. After the power was back I tried to start the cluster. Two out of 6 nodes cannot start because of this exception:

 INFO 20:47:11,003 Initializing system.local
 INFO [main] 2013-12-27 20:47:11,003 ColumnFamilyStore.java (line 251) Initializing system.local
 INFO 20:47:11,006 Opening /hadoop/disk1/cassandra/data/system/local/system-local-jb-2478 (5836 bytes)
 INFO [SSTableBatchOpen:1] 2013-12-27 20:47:11,006 SSTableReader.java (line 223) Opening /hadoop/disk1/cassandra/data/system/local/system-local-jb-2478 (5836 bytes)
 INFO 20:47:11,006 Opening /hadoop/disk4/cassandra/data/system/local/system-local-jb-2479 (144 bytes)
 INFO [SSTableBatchOpen:2] 2013-12-27 20:47:11,006 SSTableReader.java (line 223) Opening /hadoop/disk4/cassandra/data/system/local/system-local-jb-2479 (144 bytes)
ERROR 20:47:12,366 Exception encountered during startup
java.lang.IllegalArgumentException: bufferSize must be positive
        at org.apache.cassandra.io.util.RandomAccessReader.<init>(RandomAccessReader.java:67)
        at org.apache.cassandra.io.compress.CompressedRandomAccessReader.<init>(CompressedRandomAccessReader.java:76)
        at org.apache.cassandra.io.compress.CompressedRandomAccessReader.open(CompressedRandomAccessReader.java:55)
        at org.apache.cassandra.io.sstable.SSTableReader.openDataReader(SSTableReader.java:1363)
        at org.apache.cassandra.io.sstable.SSTableScanner.<init>(SSTableScanner.java:67)
        at org.apache.cassandra.io.sstable.SSTableReader.getScanner(SSTableReader.java:1147)
        at org.apache.cassandra.db.RowIteratorFactory.getIterator(RowIteratorFactory.java:69)
        at org.apache.cassandra.db.ColumnFamilyStore.getSequentialIterator(ColumnFamilyStore.java:1526)
        at org.apache.cassandra.db.ColumnFamilyStore.getRangeSlice(ColumnFamilyStore.java:1645)
        at org.apache.cassandra.db.RangeSliceCommand.executeLocally(RangeSliceCommand.java:137)
        at org.apache.cassandra.cql3.statements.SelectStatement.executeInternal(SelectStatement.java:236)
        at org.apache.cassandra.cql3.statements.SelectStatement.executeInternal(SelectStatement.java:1)
        at org.apache.cassandra.cql3.QueryProcessor.processInternal(QueryProcessor.java:255)
        at org.apache.cassandra.db.SystemKeyspace.getUnfinishedCompactions(SystemKeyspace.java:206)
        at org.apache.cassandra.service.CassandraDaemon.setup(CassandraDaemon.java:261)
        at org.apache.cassandra.service.CassandraDaemon.activate(CassandraDaemon.java:461)
        at org.apache.cassandra.service.CassandraDaemon.main(CassandraDaemon.java:504)

Collecting the logs now, will attach to the issue in a moment.