On HDP 2.4, we're using the standard Hue with security on (Hue 2.6.1-258) We have map reduce, pig, hive and spark running successfully from the command line and Kerberos. In Hue, we have also everything working fine, expect when we submit a Map Reduce job: the job is submitted and working, but when looking at it from the job history, we are getting the following error: Job job_1474295224072_0004 could not be found: <html> <head> <meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1"/> <title>Error 401 Authentication required</title> </head> <body><h2>HTTP ERROR 401</h2> <p>Problem accessing /ws/v1/history/mapreduce/jobs/job_1474295224072_0004. Reason: <pre> Authentication required</pre></p><hr /><i><small>Powered by Jetty://</small></i><br/> <br/> <br/> <br/> <br/> <br/> <br/> <br/> <br/> <br/> <br/> <br/> <br/> <br/> <br/> <br/> <br/> <br/> <br/> <br/> </body> </html> (error 401) When submitting TEZ or Spark jobs, there is no issue, we can see the job details ( it's actually going to another REST url to query for information). The job browser problem is only limited to Map Reduce jobs. When tracking down the issue, we see the following URL in the Hue logs for the job history: http_client DEBUG REST invocation: curl -X GET -H Accept: application/json 'http://my_cluster:19888/ws/v1/history/mapreduce/jobs/job_1474295224072_0004' Which is wrong. It should contain a --negotiate such as: http_client DEBUG REST invocation: curl -X GET --negotiate -H Accept: application/json 'http://my_cluster:19888/ws/v1/history/mapreduce/jobs/job_1474295224072_0004' Looking at the source code shows no indication of any kind of support for that. So my question is: is Hue shipped in HDP 2.4 capable to use Kerberos for this job browser? If not, any hints on what is wrong in our setup?
... View more
