1

i was working with Cassandra 1.2.4 probably, after restoring some key-space when i tried to query in a key-space it gave me Request did not complete within rpc_timeout

so i checked system.log & output.log under /var/log/cassandra path

i just find this exception:

Exception in thread Thread[ReadStage:42,5,main] java.lang.RuntimeException: org.apache.cassandra.io.sstable.CorruptSSTableException: java.io.EOFException

what is the reason ? and how can i get rid of rpc_timeout

thanks in advance,

Mostafa Jamareh
  • 1,389
  • 4
  • 22
  • 54

2 Answers2

1

Seems somehow your SStables are corrupted. You can try rebuilding them using nodetool's scrub [keyspace] operation.

If you can't access a specific keyspace,

> ./nodetool -u <username> -pw <password> -h <cassandra_ip> scrub <keyspace>

or if you can't access any keyspace,

> ./nodetool -u <username> -pw <password> -h <cassandra_ip> scrub
Bee
  • 12,251
  • 11
  • 46
  • 73
  • thanks @Bhathiya , but it give me this exception `Exception in thread "main" FSWriteError in /var/lib/cassandra/data/key_space/table_name/snapshots/pre-scrub-1391324805458/key_space-table_name-ic-10-Statistics.db` – Mostafa Jamareh Feb 02 '14 at 07:08
  • 1
    That is a file system permission issue. Give r/w permissions to cassandra directory in /var/lib/ – Bee Feb 02 '14 at 11:20
1

cqlsh returns rpc_timeout when any error occur in server. (remote procedure call -to server- timed out).

i think you problem was after a backup/restore and the restoring step may not perform correctly and your sstables corrupted. this may be helpful.

Community
  • 1
  • 1
Farvardin
  • 5,336
  • 5
  • 33
  • 54