Member since
04-22-2014
1218
Posts
341
Kudos Received
157
Solutions
My Accepted Solutions
| Title | Views | Posted |
|---|---|---|
| 26376 | 03-03-2020 08:12 AM | |
| 16523 | 02-28-2020 10:43 AM | |
| 4794 | 12-16-2019 12:59 PM | |
| 4508 | 11-12-2019 03:28 PM | |
| 6777 | 11-01-2019 09:01 AM |
02-07-2018
09:44 AM
@nehanirmal, The agent package includes dependencies for any packages required for CDH to run. In this case, it apepars the apache2 http server is not found anywhere. It is used for Hue's Load Balancer. You could either: (1) install it manually (2) add a repository that contains the package (3) try skipping it and see what happens. You would not be able to use the Load Balancer included with the Hue service, though. Here is a list of third party packages that are required: https://www.cloudera.com/documentation/enterprise/latest/topics/cm_ig_package_dependencies.html
... View more
02-07-2018
09:34 AM
1 Kudo
@DivyaniS, What URL are you using in the browser? What is the netstat -nap |grep 7180 output (when run on the Cloudera Manager host?) What is the error you see exactly? @iram, You need to have permission in the OS file system to view the log file. You will need sudo or root access. Also note that I said "cloudera-scm-server.log", not the agent log. I would recommend opening new threads for all of these as every issue here appears to differ and there would be no one solution. -Ben
... View more
02-06-2018
01:07 PM
Hi @MSharma, I wonder if you hit a temporary outage. I just tried and it seems that the servers are up and I can download the fiels. Can you try again now and let us know if you still have an issue? Thanks!
... View more
02-06-2018
09:40 AM
@Rajukiran, That indicates that there is something wrong in HDFS. The client could not write out blocks since there were no available DataNodes. I would check the health of your NameNode via the NameNode Web UI and verify that it shows the DataNodes are up and heartbeating. Also, try using the command line to write a file to HDFS and see if you have a similar problem.
... View more
02-06-2018
09:04 AM
@Rajukiran, Check the logs for more information: Service Monitor (/var/log/cloudera-scm-firehose/*SERVICE* file)... look at the latest one. Service Monitor initiates the canary test, so if it fails, there is likely more information or a stack trace Active NameNode (/var/log/hadoop-hdfs/*NAMENODE* file). Look at the time the failure was logged in the Service Monitor log and then go look at the NameNode log on the active namenode host at the same time. If HDFS returned an error, then its hould be logged there too.
... View more
02-05-2018
09:51 AM
1 Kudo
@PrashantAgrawal, You need to either have your DataNode HTTP Web UI Port and DataNode Transceiver Port set to privileged ports or you need to do that or configure TLS to protect the HDFS connections. If you configured Kerberos via Cloudera Manager, the wizard would have made the port changes for you.
... View more
02-01-2018
11:30 AM
1 Kudo
@hamed2005, I looked more closely at the exception and now think I understand... com.mysql.jdbc.ExportControlled.getSSLSocketFactoryDefaultOrConfigured(ExportControlled.java:428) It seems that the MySQL JDBC driver may require a password. I was testing in an environment where I wasn't using TLS for my db connection, so it didn't matter. Not sure how to get around this... I'll let you know if I come up with something
... View more
02-01-2018
10:42 AM
@NewBee22, The key error in your output is: Failed to generate and store Agent UUID When the agent starts, it will check /var/lib/cloudera-scm-agent/ to see if there is a "uuid" file. If there isn't, the agent will create the file with a unique uuid value. It seems this operation fails. Since the agent runs as root, it is possible there is something else preventing file creation I suggest looking at the file path. Run: # ls -lrt /var/lib/cloudera-scm-agent Since we see that uuid_file = open(uuid_path, 'w') fails I believe that it is most likely that "cloudera-scm-agent" does not exist. If it doesn't, create /var/lib/cloudera-scm-agent and try restarting the agent.
... View more
01-31-2018
01:52 PM
@hamed2005, The password on the file does not matter for read operations. My point was that Cloudera Manager does not need to know the password of the truststore to access what it needs. I tested on CM 5.13.1 with the following in /etc/default/cloudera-scm-server and was able to auth to LDAPS fine: export CMF_JAVA_OPTS="-Xmx2G -XX:MaxPermSize=256m -XX:+HeapDumpOnOutOfMemoryError -XX:HeapDumpPath=/tmp -Djavax.net.ssl.trustStore=/usr/java/jdk1.8.0_60/jre/lib/security/jssecacerts" What does your "export" line look like?
... View more
01-31-2018
10:01 AM
Hello @Platform, Since this is a community, you get the benefit of many people looking at your questions or issues. We prefer to keep our discussions here public unless there are sensitive data involved. Please start a new thread and explain what you have tried and what the result was and I and many others will be happy to help you out. Thanks, Ben
... View more