Hi @Spyros, @balance002, @Vivian, @Romainr, I have been having the same problems and have finally now arrived at a solution which I thought I should share here: My setup: I have Hadoop/HDFS/Yarn running in one docker container and Hue running in a separate docker container (both on the same machine). However, the following should work with any other setup as well, e.g. if you have Hadoop and Hue installed in one container or installed directly on your machine. It is important, however, that you have a running instance of Hadoop/HDFS. Step 1: Configuration of your HDFS instance Add the following property to hdfs-site.xml (in my installation located in /usr/local/hadoop/etc/hadoop/hdfs-site.xml): <property> <name>dfs.webhdfs.enabled</name> <value>true</value> </property> Add the following properties to core-site.xml (in my installation located in /usr/local/hadoop/etc/hadoop/core-site.xml): <property> <name>hadoop.proxyuser.hue.hosts</name> <value>*</value> </property> <property> <name>hadoop.proxyuser.hue.groups</name> <value>*</value> </property> Restart HDFS and Yarn: cd to the bin directory where the start/stop scripts are located (in my case it's /usr/local/hadoop/sbin/) and call: ./stop-yarn.sh & ./stop-dfs.sh ./start-dfs.sh & ./start-yarn.sh Step 2: Add new directory and adjust access rights in HDFS Prerequisite: You can call the hdfs command directly on your HDFS host. If not, add the directory containing the respective binary to your PATH (on my machine: /usr/local/hadoop/bin). First, change the owner of / to hdfs: hdfs dfs -chown hdfs:hdfs / Then create a dir for the user hdfs (This user is used by Hue. It can be configured in the hue.ini via default_hdfs_superuser) and adjust ownership: hdfs dfs -mkdir /user/hdfs hdfs dfs -chown hdfs:hdfs /user/hdfs Step 3: Configuration of your Hue instance Find your hui.ini (instructions here: http://gethue.com/how-to-configure-hue-in-your-hadoop-cluster/), in my case it's /hue/desktop/conf/pseudo-distributed.ini. Find the line webhdfs_url=http://localhost:50070/webhdfs/v1 and uncomment it, in case it is commented out. Then, if your HDFS instance is not running in the same container or host, change localhost to the IP adress of where your HDFS instance is running. Step 4 (only if Hue and HDFS are not on the same container/host): Add HDFS host to /etc/hosts This step depends on your setup: If you have Hue and HDFS running in two containers, you need to link them, i.e. when calling docker run on your Hue container, add the following parameters (I know that --link is deprecated, but it works and is easier to explain): --link <HDFS-container-ID>:<HDFS-container-ID/Hostname> If you are not using containers, simply add the following line to the /etc/hosts file of your Hue host: <IP of HDFS host> <hostname of HDFS host> Here is why you need this: Hue will first adress the HDFS host via its IP (as you configured it in hue.ini). However, it will later on, for some requests, use the hostname instead of the IP, which can only be resolved by the Hue host if it's in the /etc/hosts file. Then restart Hue (in my case I simply stop and start the container) and your done 🙂 ... View more

@George Meltser Unfortunately not. I have given up and I am now building the Hadoop stack manually without Ambari. ... View more

Thank you @Namit Maheshwari. I did improve the memory settings as mentioned in your referenced article. However, it is not possible to restart all affected services after a parameter change. E.g. the Hive Metastore service always aborts its restart. Stopping and restarting the container does not help either. If I run the query anyway (without restarting), the error is this: java.sql.SQLException: Error while processing statement: FAILED: Execution Error, return code 2 from org.apache.hadoop.hive.ql.exec.tez.TezTask. Vertex failed, vertexName=Map 1, vertexId=vertex_1494585560171_0001_1_00, diagnostics=[Task failed, taskId=task_1494585560171_0001_1_00_000000, diagnostics=[TaskAttempt 0 failed, info=[Error: Failure while running task:java.lang.IllegalArgumentException: tez.runtime.io.sort.mb 2568 should be larger than 0 and should be less than the available task memory (MB):192 at com.google.common.base.Preconditions.checkArgument(Preconditions.java:92) at org.apache.tez.runtime.library.common.sort.impl.ExternalSorter.getInitialMemoryRequirement(ExternalSorter.java:327) at org.apache.tez.runtime.library.output.OrderedPartitionedKVOutput.initialize(OrderedPartitionedKVOutput.java:93) ... View more

Thank you all for your replies. I did improve the memory settings (according to https://community.hortonworks.com/articles/14309/demystify-tez-tuning-step-by-step.html). However, it is not possible to restart all affected services after a parameter change. E.g. the Hive Metastore service always aborts its restart. Stopping and restarting the container does not help either. ... View more

Hi Graham, thanks for your reply. Generally SQL queries work. However, of the two queries you posted only the first one completed (but empty result). The second one (analyze table truck_mileage compute statistics for columns;) gave me the same java heap space exception. And no, the sandbox is not using its maximum resources, not even close. I tried playing around with the container sizes, but that lead to the queries not returning at all anymore (i.e. no result, no crash, just keep running). ... View more