Support Questions
Find answers, ask questions, and share your expertise
Announcements
Alert: Welcome to the Unified Cloudera Community. Former HCC members be sure to read and learn how to activate your account here.

Cloudera Installation Path B Fails -

Highlighted

Cloudera Installation Path B Fails -

Explorer
  • Parent Command: First Run ( )

Command Details: Deploy Client Configuration

Deploy Client Configuration - Failed to deploy client configuration

Execute command deployClientConfig on service HDFS - Failed 

Completed only 0/5 steps. First failure: Client configuration (id=1) on host <x.x.x.x> (id=4) exited with 1 and expected 0.

Execute command deployClientConfig on service Solr - Failed

Completed only 0/1 steps. First failure: Client configuration (id=2) on host<x.x.x.x> (id=1) exited with 1 and expected 0.

..... 

 

I applied the   cloudera-scm ALL=(ALL) NOPASSWD: using visudo on a Centos6.6 but still the installation fails,

usermod -a -G sudo cloudera-scm gave 

usermod: group 'sudo' does not exist

 

In the manager server log, I see

2016-04-05 11:26:40,518 ERROR main:org.hibernate.engine.jdbc.spi.SqlExceptionHelper: ORA-00942: table or view does not exist

 

In the manager agent log, I see

SError: [Errno 13] Permission denied: '/var/lib/zookeeper'
[05/Apr/2016 11:49:14 +0000] 4067 MainThread agent ERROR Could not evaluate resource {u'io': None, u'named_cpu': None, u'tcp_listen': None, u'dynamic': False, u'rlimits': None, u'file': None, u'memory': None, u'directory': {u'path': u'/var/lib/zookeeper', u'bytes_free_warning_threshhold_bytes': 0, u'group': u'cloudera-scm', u'user': u'cloudera-scm', u'mode': 493}, u'cpu': None}
Traceback (most recent call last):

 

In the Agent log, i see the following errors

Failed to connect to previous supervisor.

[05/Apr/2016 11:39:52 +0000] 4067 MainThread agent ERROR Failed to connect to previous supervisor.
Traceback (most recent call last):
File "/usr/lib64/cmf/agent/src/cmf/agent.py", line 1529, in find_or_start_supervisor
self.configure_supervisor_clients()
File "/usr/lib64/cmf/agent/src/cmf/agent.py", line 1768, in configure_supervisor_clients

 

 

[05/Apr/2016 11:42:42 +0000] 4067 Monitor-HostMonitor throttling_logger ERROR Timeout with args ['/usr/java/jdk1.7.0_67-cloudera/bin/java', '-classpath', '/usr/share/cmf/lib/agent-5.4.7.jar', 'com.cloudera.cmon.agent.DnsTest']
None
[05/Apr/2016 11:42:44 +0000] 4067 Monitor-HostMonitor throttling_logger ERROR Failed to collect java-based DNS names
Traceback (most recent call last):
File "/usr/lib64/cmf/agent/src/cmf/monitor/host/dns_names.py", line 64, in collect
result, stdout, stderr = self._subprocess_with_timeout(args, self._poll_timeout)

 

 

[05/Apr/2016 11:45:23 +0000] 4067 MainThread parcel ERROR Failed to activate alternatives for parcel CDH-5.4.7-1.cdh5.4.7.p0.3: 1
[05/Apr/2016 11:45:23 +0000] 4067 MainThread parcel INFO Executing command ['/usr/lib64/cmf/service/common/alternatives.sh', 'activate', 'solr-conf', '/etc/solr/conf', 'etc/solr/conf.dist', '10', 'True']
[05/Apr/2016 11:45:23 +0000] 4067 MainThread parcel ERROR Failed to activate alternatives for parcel CDH-5.4.7-1.cdh5.4.7.p0.3: 1
[05/Apr/2016 11:45:23 +0000] 4067 MainThread parcel INFO Executing command ['/usr/lib64/cmf/service/common/alternatives.sh', 'activate', 'hive-hcatalog-conf', '/etc/hive-hcatalog/conf', 'etc/hive-hcatalog/conf.dist', '10', 'True']

 

I am installing on AWS using the Centos 6.6 AMI ami-f35e56c3. is there anything else that needs to be done?

 

4 REPLIES 4

Re: Cloudera Installation Path B Fails -

Explorer

Can somebody please help on this issue!!

 

I am stuck with this for the past 2 weeks, not much information from the cloudera communities as well.

 

I have installed various version of cloudera more than 10 times but nothing worked

 

 

Re: Cloudera Installation Path B Fails -

Cloudera Employee

This appears to be a permission issue. One thing I noticed is that you are missing the "ALL" at the end of you visudo entry.

For example, in my local instance I use:

myuser ALL=(ALL) NOPASSWD:ALL

Also, make sure you apply this across all nodes in your cluster.

Re: Cloudera Installation Path B Fails -

Super Guru

Hello mukundrv,

 

I'm sorry to hear you are having trouble with the installations.

 

From what I gather of your description, installation failed at the "Deploy Client Configuration" step.  If that is the case, then I believe installation was, for the most part, successful and you should be able to exit the wizard and manage the cluster and services.  You could open a new browser tab to access Cloudera Manager.

 

Then we can take a closer look at why deploying the client configuration is failing.

 

That said, you outlined a number of errors, but it is not easy to tell which are relevant to the issue you present:

 

(1)

 

2016-04-05 11:26:40,518 ERROR main:org.hibernate.engine.jdbc.spi.SqlExceptionHelper: ORA-00942: table or view does not exist

 

When did that occur?  Does it happen if you try clicking the Drop-down beside the cluster name (in the Cloudera Manager Home Screen) and choose "Deploy Client Configuration?"  To tell more, we'd need to review the entire stack trace as well.

 

(2)

 

SError: [Errno 13] Permission denied: '/var/lib/zookeeper'

 

That error is a bit odd.  If it is the only "Permission denied" error, then it indicates there may be a /var/lib/zookeeper directory that could not be managed by the agent.  Try restarting that agent with service cloudera-scm-agent restart and see if there are further complaints.

 

(3)

 

[05/Apr/2016 11:39:52 +0000] 4067 MainThread agent ERROR Failed to connect to previous supervisor.
Traceback (most recent call last):
File "/usr/lib64/cmf/agent/src/cmf/agent.py", line 1529, in find_or_start_supervisor

 

Don't worry about that one.  This happens every time the agent starts and there is no supervisor started.  When the agent is started, it tries to connect to a supervisor.  If there is none, the agent starts the supervisor and then connects to it.  If you look at the logs after the above exception, you should see lines that support my description.

 

(4)

 

[05/Apr/2016 11:42:44 +0000] 4067 Monitor-HostMonitor throttling_logger ERROR Failed to collect java-based DNS names

 

This is non-fatal, but indicates there may be problems with DNS on that host.  If you look just before or after the exception, you will find the command line arguments used to run the small java DNS checker.  You can try using those to see if the issue may be transient.   Also, to debug this, we would need to see the full stack trace and perhaps 5 lines before and after the exception.

 

(5)

 

[05/Apr/2016 11:45:23 +0000] 4067 MainThread parcel ERROR Failed to activate alternatives for parcel CDH-5.4.7-1.cdh5.4.7.p0.3: 1

 

This looks pretty severe since it means that the attempt to activate the parcel on the host from which you obtained the error failed.

My guess is that there were some permissions errors that made updating alternatives impossible.  Really, in order to debug this, we would need to see a full agent log from one or more of the hosts.

 

That said, I think there is evidence here that the sudo permissions may not be adequate and some verification that cloudera-scm does actually have the necessary privileges should be done.  Also make sure you have disabled selinux as that can cause issues in odd places.  It can be reenabled later if you have it enabled and need it.

 

I'm not sure of the best ways to verify sudo for cloudera-scm.

 

Ben

Re: Cloudera Installation Path B Fails -

Explorer

Thanks elias, bgooley, for your response!!

 

All the logs which were collected above were after the deployment configuration failed in both the servers and the agentse

During my efforts, I was trying the single user mode installation, which was failing on every occasion. 

When I tried a non single user mode installation, it worked. 

 

Would appreciate if there could be more documentation on the list of folders to be created for the single user mode installation.

Don't have an account?
Coming from Hortonworks? Activate your account here