giraph-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Suijian Zhou <suijian.z...@gmail.com>
Subject computation stucks in giraph.
Date Wed, 02 Apr 2014 15:27:30 GMT
Hi,
  When I run a giraph program, it hangs there forever with output like:
14/04/02 10:16:43 INFO zookeeper.ClientCnxn: Opening socket connection to
server compute-0-16.local/10.1.255.238:22181. Will not attempt to
authenticate using SASL (unknown error)
14/04/02 10:16:43 INFO zookeeper.ClientCnxn: Socket connection established
to compute-0-16.local/10.1.255.238:22181, initiating session
14/04/02 10:16:43 INFO zookeeper.ClientCnxn: Session establishment complete
on server compute-0-16.local/10.1.255.238:22181, sessionid =
0x1452302cc870005, negotiated timeout = 600000
14/04/02 10:16:43 INFO job.JobProgressTracker: Data from 4 workers -
Compute superstep 0: 1000 out of 1000 vertices computed; 16 out of 16
partitions computed; min free memory on worker 0 - 179.86MB, average
202.77MB
14/04/02 10:16:44 INFO mapred.JobClient:  map 100% reduce 0%
14/04/02 10:16:48 INFO job.JobProgressTracker: Data from 4 workers -
Compute superstep 0: 1000 out of 1000 vertices computed; 16 out of 16
partitions computed; min free memory on worker 0 - 179.86MB, average
201.15MB
14/04/02 10:16:53 INFO job.JobProgressTracker: Data from 4 workers -
Compute superstep 0: 1000 out of 1000 vertices computed; 16 out of 16
partitions computed; min free memory on worker 0 - 179.86MB, average
201.15MB
14/04/02 10:16:58 INFO job.JobProgressTracker: Data from 4 workers -
Compute superstep 0: 1000 out of 1000 vertices computed; 16 out of 16
partitions computed; min free memory on worker 0 - 179.86MB, average
201.15MB
14/04/02 10:17:03 INFO job.JobProgressTracker: Data from 4 workers -
Compute superstep 0: 1000 out of 1000 vertices computed; 16 out of 16
partitions computed; min free memory on worker 3 - 174.34MB, average 207.9MB
14/04/02 10:17:08 INFO job.JobProgressTracker: Data from 4 workers -
Compute superstep 0: 1000 out of 1000 vertices computed; 16 out of 16
partitions computed; min free memory on worker 3 - 174.34MB, average 207.9MB
14/04/02 10:17:13 INFO job.JobProgressTracker: Data from 4 workers -
Compute superstep 0: 1000 out of 1000 vertices computed; 16 out of 16
partitions computed; min free memory on worker 3 - 174.34MB, average 207.9MB

The input is only a common graph with 1000 lines and 800kB size. But the
program runs without problem for small graph and returns expected results.
What are the possible reasons? Thanks!

  Best Regards,
  Suijian

Mime
View raw message