Hi all, I recently tried Giraph.
It compiled fine, but when I did the unit test (i.e. `mvn test`) it failed. I did the test several times with different settings. See below for details. My question is: - Is it normal for Giraph? Is the unit test supposed to fail for now? - If it's not normal, what do you think is wrong? My environment: - Ubuntu Server 14.04 64-bit - Dynamic address by DHCP - Followed the "Quick Start" guide https://giraph.apache.org/quick_start.html except I used "localhost" for the node hostname. -- Installed openjdk-7-jdk, git, maven by apt-get -- Installed hadoop-0.20.203.0rc1 as instructed Test settings: (1) 1.1.0-SNAPSHOT (Commit ID: b218d72cedc52467e691c6002e596e482d8583e4) with LocalJobRunner (2) 1.1.0-SNAPSHOT with the running Hadoop instance (3) Tag "release-1.0.0-RC3" with LocalJobRunner (4) Tag "release-1.0.0-RC3" with the running Hadoop instance Cases (1) and (3) are run by the following commands. $ mvn clean $ mvn package -Phadoop_0.20.203 -DskipTests $ mvn test -Phadoop_0.20.203 Cases (2) and (4) are run by the following commands. $ mvn clean $ mvn package -Phadoop_0.20.203 -DskipTests $ mvn test -Phadoop_0.20.203 -Dprop.mapred.job.tracker=localhost:54311 Test results: (1) Failed at "Rexster I/O Format". At some point, the test endlessly tried to connect to ZooKeeper, repeating the following log messages. I had to terminate the test by Ctrl-C. 14/06/26 18:10:27 INFO zookeeper.ClientCnxn: Opening socket connection to server localhost/127.0.0.1:22182. Will not attempt to authenticate using SASL (unknown error) 14/06/26 18:10:27 WARN zookeeper.ClientCnxn: Session 0x146d76e16750002 for server null, unexpected error, closing socket connection and attempting reconnect java.net.ConnectException: Connection refused at sun.nio.ch.SocketChannelImpl.checkConnect(Native Method) at sun.nio.ch.SocketChannelImpl.finishConnect(SocketChannelImpl.java:739) at org.apache.zookeeper.ClientCnxnSocketNIO.doTransport(ClientCnxnSocketNIO.java:350) at org.apache.zookeeper.ClientCnxn$SendThread.run(ClientCnxn.java:1068) (2) Failed at "Examples". Hadoop was stuck at "testBspFail" job. Logs like the following line were printed every 5 seconds for 10 minutes. After that, the test endlessly tried to connect to ZooKeeper, just like the case (1). I killed the test process and the Hadoop job. 14/06/26 18:40:30 INFO job.JobProgressTracker: Data from 2 workers - Compute superstep 20: 10 out of 10 vertices computed; 6 out of 6 partitions computed; min free memory on worker 1 - 137.42MB, average 149.12MB (3) Failed at "HBase I/O" in testHBaseInputOutput (org.apache.giraph.io.hbase.TestHBaseRootMarkerVertextFormat). Before it reported failure, it blocked for 10 minutes. (4) Failed at "Core" in testContinue (org.apache.giraph.io.TestJsonBase64Format) . Hadoop was stuck at the second "testContinue" job. After 10 minutes, the test went on and reported the failure. The "testContinue" map task was aborted with the following error. java.io.FileNotFoundException: /tmp/_giraphTests/testContinue/_logs (Is a directory) Thanks in advance. ------------------------------------ Toshio Ito
