Member since
03-09-2018
12
Posts
1
Kudos Received
0
Solutions
02-17-2019
06:11 PM
In fact the problem is related to the java Orc api when parallelism is activated (multi-thread) I use Flink and when I set a parallelism > 1 on the Sink that generates Orc files, I face to this issue: data are unreadable. I've seen some tickets about this issue like this one: https://jira.apache.org/jira/browse/ORC-361 At the moment I use a parallelism of 1 but I have to fix this issue in order to scale my ingest pipeline. All help is welcome. Thx
... View more
02-11-2019
02:34 PM
Hello, I face to one error when I try to read my Orc files from Hive (external table) or Pig or with hive --orcfiledump .. These files are generated with Flink using the Orc Java API with Vectorize column. If I create these files locally (/tmp/...), push them to hdfs, then I can read the content of these files from Pig or with the use of External table in Hive. If I change the path and use hdfs directly, then I face to this error : Failure while running task:java.lang.NullPointerException
at org.apache.hadoop.hive.ql.io.orc.RunLengthIntegerReaderV2.readValues(RunLengthIntegerReaderV2.java:54)
at org.apache.hadoop.hive.ql.io.orc.RunLengthIntegerReaderV2.next(RunLengthIntegerReaderV2.java:302)
at org.apache.hadoop.hive.ql.io.orc.TreeReaderFactory$TimestampTreeReader.next(TreeReaderFactory.java:1105)
at org.apache.hadoop.hive.ql.io.orc.TreeReaderFactory$StructTreeReader.next(TreeReaderFactory.java:2079)
at org.apache.hadoop.hive.ql.io.orc.RecordReaderImpl.next(RecordReaderImpl.java:1082)
at org.apache.hadoop.hive.ql.io.orc.OrcNewInputFormat$OrcRecordReader.nextKeyValue(OrcNewInputFormat.java:108) And the same if I get these files locally.
... View more
Labels:
07-26-2018
04:41 PM
Hello,
I've just read in detail the Knox dev guide. But I face to an issue with logsearch on my HDP cluster.
This is a HDP 2.6.5 with Knox 0.12.0
I've created some XML files. These files are in PJs.
But I face to some 403 requests. In fact, some urls have been "urlencoded" twice: Rewrote URL: http://XXX:80/gateway/default/logsearch/api/v1/service/logs/histogram?page=0&pageSize=999999999&startIndex=0&q=*%3A*... direction: IN
via implicit rule: LOGSEARCH/logsearch/inbound to URL: http://XXX:61888/api/v1/service/logs/histogram?q=*%253A*&startIndex=0... Thus, we've got the string "q=%3A" replaced by "q=%253A" How can I resolve this issue ?
I've written a custom dispatch class and reference it in the service.xml but my class seems to be ignored.
<dispatch classname="org.apache.hadoop.gateway.logsearch.LogsearchDispatch"/>
I've put my Jar that contains this class in the ext directory.Thus, it should be in the classpath.
Thanks in advance
Regards, David
... View more
07-25-2018
02:38 PM
rewrite.xmlservice.xmlHello, I face to some issues with the integration of logsearch with Knox.
My cluster is an HDP 2.6.5. The Knox version with HDP 2.6.5 is 0.12.0
I've defined the LOGSEARCH topology with custom Xml files but I face to some internal rewrite issues: https://XXXX/gateway/default/logsearch/api/v1/service/logs/histogram?page=0&pageSize=999999999&startIndex=0&q=*%3A* http://XXXX:61888/api/v1/service/logs/histogram?q=*%253A*.
q=*%3A* replaced by q=*%253A* => double urlencode. Thus logsearch can't handle this kind of request.
I'm trying to define a custom dispatch class. The documentation explains a little how to do that but it's not sufficient. I'd like to find a concrete example. Where can I find a concrete example of Dispatch class embedded that I can put in a dedicated jar (then I can copy this jar in the ext directory) ?
Or how can I handle handle this kind of double encode ? Thanks in advance Regards, David
... View more
03-25-2018
10:43 PM
1 Kudo
Related post: https://community.hortonworks.com/questions/176408/hdp-264-problem-during-application-of-kerberos.html Hello, I faced to some issues during the creation of one kerberized cluster with Ambari blueprint. All the issues are resolved. This post is only a feedback (for sharing) Here after the different issues
- (Minor) Unset variable {{policymgr_mgr_url}} in the blueprint that
causes this error in /var/lib/ambari-agent/data/output-XX.txt:
"..returned 3. curl: (3) [globbing] nested braces not supported at pos
2\n",). Sleeping for 8 sec(s)". This error increases significantly the time to create cluster (lot of retries) => resolved by setting the parameter in the blueprint. {
"ranger-hdfs-security" : {
"properties_attributes" : { },
"properties" : {
"ranger.plugin.hdfs.policy.source.impl" : "org.apache.ranger.admin.client.RangerAdminRESTClient",
"ranger.plugin.hdfs.policy.rest.url" : "http://XXX:6080"
}
}
}
- (Minor) Install of KMS failed due to an unset db password. => resolved by defined this KMS user/password in blueprint
{
"kms-properties" : {
"properties_attributes" : { },
"properties" : {
"DB_FLAVOR" : "MYSQL",
...
"db_user" : "XXX",
"db_password" : "XXXX"
}
}
}
- (Major) The main problem is the fact that Kerberos
keytabs are not been copied by Ambari onto one server of my cluster.
This server hosts some HDP components especially a journalnode. Without
keytabs, the journalnode cannot start.
Thus, format of the Namenode when the cluster is starting fails. This
process is in background and logs are in a dedicated logs file. That´s
why it´s not really obvious.
Workaround: when, I force manually the "Regenerate keytabs" through
Ambari UI (Admin/Kerberos), Keytabs are well generated and copied onto
this server. Then, the journalnode is able to start and then Namenodes. => resolved. As soon as I've resolved the other problems (KMS, policymgr_mgr_url...), this problem has been resolved automatically. To my point of view, the error in the install of the other components blocked the propagation of the keytabs. Not really obvious to identify. Regards, David
... View more
03-21-2018
11:06 PM
One HCC thread related to this kind of error: https://community.hortonworks.com/questions/84101/failed-to-start-hive-server2-illegal-character-in.html
... View more
03-21-2018
11:02 PM
I pursue in the resolution of my problem and my understanding.. When I activate HDFS Audit in Ranger Configs and restart a Namenode to apply config, I obtain these errors in the log file '
/var/lib/ambari-agent/data/output-XX.txt':
call returned (3, '') 2018-03-21
23:44:37,745 - Will retry 7 time(s), caught exception: (u"Execution of
'curl --location-trusted -k --negotiate -u : -b
/var/lib/ambari-agent/tmp/cookies/7768 3cd0-2605-4b42-a750-509ecc713ba9
-c
/var/lib/ambari-agent/tmp/cookies/77683cd0-2605-4b42-a750-509ecc713ba9
-w '%{http_code}' '{{policymgr_mgr_url}}/login.jsp' --connect -timeout
10 --max-time 12 -o /dev/null 1>/tmp/tmp21P7Ww 2>/tmp/tmpfrrHEy'
returned 3. curl: (3) [globbing] nested braces not supported at pos
2\n",). Sleeping for 8 sec(s)
... View more
03-21-2018
04:19 PM
Hello, I'm still trying to deploy a "kerberized" cluster HDP with Ambari blueprint. With the Ranger service but with all plugins desactivated, it's ok. A soon as I try to activate plugins, the cluster never starts (Namenodes are still stopped) : "ranger-hdfs-plugin-enabled" : "Yes"... A JVM related to the Namenode is started but that's all. This is the command to format Namenode: org.apache.hadoop.hdfs.server.namenode.NameNode -format -nonInteractive Then, there is no log file related to the namenode in /var/log/hadoop/hdfs/ If someone has some advices or an example of blueprint that works with Kerberos and Ranger (+ plugins) I will be eternally grateful 🙂
... View more
03-20-2018
02:30 PM
I think the problem is related to missing steps during post install but I'm not sure because the error is not really obvious. Example for Hdfs: https://docs.hortonworks.com/HDPDocuments/HDP2/HDP-2.6.4/bk_security/content/hdfs_plugin_kerberos.html Questions about that : - If these steps are required, does I have to create this 'rangerlookup' user manually and its keytab (cluster is kerberized) ? - Is it blueprint compliant ?
... View more
03-12-2018
12:24 AM
Despite the copy of these ranger-* files before application of Kerberos, I still obtain the same error.
... View more
03-09-2018
11:25 PM
Thanks for your reply It seems ok. I spawn a new cluster from scratch and give a feedback.
... View more
03-09-2018
10:11 PM
Hello, I face to a problem related to Ranger during application of Kerberos with Ambari onto my HDP cluster (2.6.4) FATAL conf.Configuration (Configuration.java:loadResource(2712)) -
error parsing conf file:/.../hadoop/2.6.4.0-91/0/xasecure-audit.xml java.io.FileNotFoundException: /.../hadoop/2.6.4.0-91/0/xasecure-audit.xml (No such file or directory) I've found a temporary work-around by desactivate audit in Ranger configuration. How can I resolve this problem in order to reactivate audit. Thanks in advance, Regards, David
... View more