RegionServer throws "not a SequenceFile", oldWALs growing, Hbase-indexer not indexing

[#1]

Hi,

At some point "test_indexer" stopped indexing new entries from the hbase.

The table has the replication scope set to 1. From the logs, the "test_indexer" and the solr server appear to be working just fine.

Looking at the RegionServer logs I see the following warning being repeated all day:

WARN	org.apache.hadoop.hbase.replication.regionserver.ReplicationSource	
Indexer_test_indexer Got: 
java.io.EOFException: hdfs://tets-node:8020/hbase/oldWALstest-node%2C60020%2C1480189429512.null0.1480189450431 not a SequenceFile
	at org.apache.hadoop.io.SequenceFile$Reader.init(SequenceFile.java:1919)
	at org.apache.hadoop.io.SequenceFile$Reader.initialize(SequenceFile.java:1878)
	at org.apache.hadoop.io.SequenceFile$Reader.<init>(SequenceFile.java:1827)
	at org.apache.hadoop.io.SequenceFile$Reader.<init>(SequenceFile.java:1841)
	at org.apache.hadoop.hbase.regionserver.wal.SequenceFileLogReader$WALReader.<init>(SequenceFileLogReader.java:70)
	at org.apache.hadoop.hbase.regionserver.wal.SequenceFileLogReader.reset(SequenceFileLogReader.java:168)
	at org.apache.hadoop.hbase.regionserver.wal.SequenceFileLogReader.initReader(SequenceFileLogReader.java:177)
	at org.apache.hadoop.hbase.regionserver.wal.ReaderBase.init(ReaderBase.java:66)
	at org.apache.hadoop.hbase.wal.WALFactory.createReader(WALFactory.java:302)
	at org.apache.hadoop.hbase.wal.WALFactory.createReader(WALFactory.java:267)
	at org.apache.hadoop.hbase.wal.WALFactory.createReader(WALFactory.java:255)
	at org.apache.hadoop.hbase.wal.WALFactory.createReader(WALFactory.java:406)
	at org.apache.hadoop.hbase.replication.regionserver.ReplicationWALReaderManager.openReader(ReplicationWALReaderManager.java:70)
	at org.apache.hadoop.hbase.replication.regionserver.ReplicationSource$ReplicationSourceWorkerThread.openReader(ReplicationSource.java:745)
	at org.apache.hadoop.hbase.replication.regionserver.ReplicationSource$ReplicationSourceWorkerThread.run(ReplicationSource.java:541)

Folder  /hbase/oldWALs is taking up 8.7Gb and slowly but noticeably increasing.

So my guess is the indexer is not marking the WAL entries as replicated and they keep accumulating in the oldWALs folder. How do I determine the cause of the issue?

Could you also explain to me what this exception means? The set-up worked just fine a day ago.

Thanks,

Gin

[#2]

More info. I see in the oldWALs folder the oldest wal file size is 0. Can it be that this is a corrupt file and the indexer cannot get past it to process the rest of WAL records?

[#3 · Accepted solution]

Yes, the 0b-sized old WALs were the culprit.

The hbase-indexer could not get past the corrupt files no matter what. Removing them was not an option as the hbase-indexer was expecting for these files.

I did a dirty trick and copy-pasted the contents from other old WALs.The source WALs, judging from the size and contents, looked like empty carriers.

That did the trick and hbase-indexer finally consumed all of the old WALs.

[#4]

Thx, that was interesting to know !

[#5]

I am having the same error but I didn't understand the solution, can you please explain it. Thank you.