AnsweredAssumed Answered

write failures

Question asked by hastings on Jul 22, 2011
Latest reply on Jul 22, 2011 by srivas
Suppose I am updating a block with 100 bytes. Suppose write is successful at replicas A and B. But at replica C say the replica crashed after writing 99 bytes.  How do you fix the offending replica? I guess this question applies to HDFS too but I guess no one uses it for a random write workload. How does MapR handle this situation?