Member since
07-31-2013
1924
Posts
461
Kudos Received
311
Solutions
My Accepted Solutions
Title | Views | Posted |
---|---|---|
1267 | 07-09-2019 12:53 AM | |
6575 | 06-23-2019 08:37 PM | |
7160 | 06-18-2019 11:28 PM | |
7470 | 05-23-2019 08:46 PM | |
2731 | 05-20-2019 01:14 AM |
08-15-2016
08:55 AM
Assuming you want to set this on Brokers, you may append it in the syntax shown below into the UI field of CM -> Kafka -> Configuration -> "Kafka Broker Advanced Configuration Snippet (Safety Valve) for kafka.properties": num.network.threads=8 For understanding these fields better, please read our Custom Configuration CM feature documentation at http://www.cloudera.com/documentation/enterprise/latest/topics/cm_mc_config_snippet.html
... View more
08-15-2016
08:41 AM
1 Kudo
There's no real significance to it, its just the way the directory detection logic works today. The double slashes on Linux get interpreted as a single one, and does not induce the behaviour in any way (same within Java's CLASSPATH parser, for which this entry is created) Here's where the classpath entry is being built: /usr/lib/hadoop/libexec/hadoop-config.sh (you can vi this file on your installation) Sourcing the variables that are being appended to CLASSPATH ('/*' is being appended to whose values) will reveal further that some of them come from the file /usr/lib/hadoop/libexec/hadoop-layout.sh (you can vi this as well) wherein they are set to './'. As a result the concatenation operations on these strings produce the logic: "/usr/lib/hadoop" (base) + "/" (separator) + "./" (specific sub-dir path) + "/*" (wildcard suffix) The output of such a thing would thereby appear as "/usr/lib/hadoop/.//*". Similar things apply to the non wildcard question, where we look to just add the directory itself onto the classpath. Understanding-wise this is no different in evaluation than the reduced form of: /etc/hadoop/conf:/usr/lib/hadoop/lib/*:/usr/lib/hadoop/*:/usr/lib/hadoop-hdfs/:/usr/lib/hadoop-hdfs/lib/*:/usr/lib/hadoop-hdfs/*:/usr/lib/hadoop-yarn/lib/*:/usr/lib/hadoop-yarn/*:/usr/lib/hadoop-mapreduce/lib/*:/usr/lib/hadoop-mapreduce/* Hope this helps in reducing the confusion.
... View more
08-15-2016
08:16 AM
1 Kudo
These fixes will appear in CDH 5.8.2 onwards for 5.8.x series. They were pulled as bug-fixes into the branch after 5.8.0's cut. 5.7.2 has already seen the day, but 5.8.2 will arrive later. Our current release schedules are parallel for each minor version level, which would explain this observance.
... View more
08-15-2016
08:05 AM
This is standard behaviour currently. The retries of 3 times to reach an MR2 AM is controlled via "yarn.app.mapreduce.client-am.ipc.max-retries". If the client cannot reach the MR2 AM, which is the definitive source of truth during application execution, then it falls back to asking the RM which may answer the state and the redirection if the application has completed.
... View more
08-15-2016
02:30 AM
Thanks for testing it out. Can you retry the CM command again? Perhaps it was a transient problem during some library changes. Also, if the command still fails with the same error, try the full actual command CM is trying to run too: kadmin -k -t test-admin.keytab -p test/admin@ACME.COM -r ACME.COM -q 'addprinc -maxrenewlife "432000 sec" -randkey kafka/sb-node1.example.com@ACME.COM'
... View more
08-15-2016
01:29 AM
Thank you for checking and reporting back, it appears something is very strange about your Ubuntu environment. In my test environment with a 14.04 VM install, these commands appear to work normally. Can you verify if they work the same in yours (below was executed on the KDC host, for my realm name of UBUNTU.VM): ~> kadmin.local kadmin: addprinc test/admin … kadmin: xst -k test-admin.keytab -norandkey test/admin kadmin: quit ~> ~> kadmin -k -t test-admin.keytab -p test/admin@UBUNTU.VM -r UBUNTU.VM kadmin: listprincs … kadmin: quit ~> I'm particularly interested to know if both kadmin.local and kadmin (with an admin keytab) commands work for you normally outside of Cloudera Manager which is merely also invoking a similar command via a shell script.
... View more
08-15-2016
12:33 AM
Could you also post your version numbers for the below package names? libkrb5support0 [ideal: 1.12+dfsg-2ubuntu5.2] libc6 [ideal: 2.19-0ubuntu6.9]
... View more
08-14-2016
07:28 PM
You need just one of those jars in the directory, not all of them. Just use one, the one named sqljdbc41.jar (The 4.1 JDBC spec is the latest version supported with JDK7). Remove everything else (especially remove sqljdbc42.jar which won't work unless you upgrade to JDK8) to avoid this problem.
... View more
08-14-2016
04:01 PM
1 Kudo
Try stopping all roles and restarting your CM agent on just this host and ensure you do not use the init.d script directly, and that instead you use the recommended service command approach: ~> service cloudera-scm-agent restart (i.e., never do this: "~> /etc/init.d/cloudera-scm-agent restart")
... View more
08-14-2016
03:58 PM
Are you running a modified Ubuntu 14.04 with non-standard repositories? The kadmin command (from Ubuntu's packages of krb5-kdc and krb5-admin-server) we are invoking as the default PATH provides us is failing at a lower level library incompatibility error. What are your versions of these packages? They should ideally be "1.12+dfsg-2ubuntu5.2", is that matching with what you have installed? And also, did you attempt an upgrade of any other packages (such as libc, kernel, etc.) from 14.04 to a later version?
... View more