Ambari-2.6 automatic installation CLUSTER INSTALL WIZARD stuck

[#1]

Hi all,

I am confused to call it as error/issue.

While installing automatic Ambari 2.6 with HDP 2.6.3 on 4 node cluster in CLUSTER INSTALL WIZARD all went fine till i reached Review step and clicked on deploy.

The installation is stuck on step 8 of CLUSTER INSTALL WIZARD review and Deploy button faded.

Steps Performed:

1) Ambari-server and Agent restart on all 4 node for 5 times.

2) Step 0 to step 8 and step 8 to step 0 for 5 times.

3) Tested on Chrome, IE and Mozilla same steps for 3 times.

is that any thing i missed to perform please let me know?

[#2]

Found this in ambari-server.log :

21 Nov 2017 12:55:02,177 INFO [pool-18-thread-1] MetricsServiceImpl:65 - Attempting to initialize metrics sink 21 Nov 2017 12:55:02,178 INFO [pool-18-thread-1] MetricsServiceImpl:81 - ********* Configuring Metric Sink ********** 21 Nov 2017 12:55:02,178 INFO [pool-18-thread-1] AmbariMetricSinkImpl:95 - No clusters configured. 21 Nov 2017 12:55:12,117 ERROR [alert-event-bus-1] AlertReceivedListener:480 - Unable to process alert ambari_agent_version_select for an invalid cluster named abcdefg

ambari-audit.log:

2017-11-21T12:54:37.162+0530, User(admin), RemoteIp(Desktop-IP), RequestType(POST), url(http://10.128.20.10:8080/api/v1/version_definitions?dry_run=true), ResultStatus(201 Created) 2017-11-21T12:55:17.758+0530, User(admin), RemoteIp(Desktop-IP), Operation(Repository version removal), RequestType(DELETE), url(http://ambari-server_fqdn:8080/api/v1/stacks/HDP/versions/2.6/repository_versions/1), ResultStatus(500 Internal Server Error), Reason(org.apache.ambari.server.controller.spi.SystemException: Repository version can't be deleted as it is used by the following hosts: CURRENT on node03-FQDN), Stack(HDP), Stack version(2.6), Repo version ID(1)

[#3]

DO you have cluster named abcdefg ? Run below command and check

curl -U <username>:<password> http://<ambari-host>:<port>/api/v1/clusters

[#4 · Accepted solution]

@Rohit Sharma

1. Sometimes renaming a cluster might cause the following kind of error message (OR) it is also possible that somehow the cluster_id and cluster_name are not updated properly inside the database (due to some temporary failure/error during cluster creation / moving step backwards or forward during cluster creation)

12:55:12,117 ERROR [alert-event-bus-1] AlertReceivedListener:480 - Unable to process alert ambari_agent_version_select for an invalid cluster named abcdefg

.

2. We also see that there is an indication of "500 Internal Server Error", which happens when something goes wrong inside the ambari server. So i guess you should see a complete detailed stacktrace in your "/var/log/ambari-server/ambari-server.log" file around "2017-11-21T12:55:17.758" timestamp as we see the following message in your "ambari-audit.log"

2017-11-21T12:55:17.758+0530, User(admin), RemoteIp(Desktop-IP), 
Operation(Repository version removal), RequestType(DELETE), 
url(http://ambari-server_fqdn:8080/api/v1/stacks/HDP/versions/2.6/repository_versions/1),
 ResultStatus(500 Internal Server Error),

.

Workaround:
As this is a fresh cluster that you are trying to setup so it will be quick if you can reset the ambari DB by running the following command (which will wipe up the incomplete cluster details from DB) and then freshly try to setup the ambari cluster. else the other approach will be to look at the DB dump and fix the entries inside the few tables like "clusters" , "host_version" ...etc to make sure that it is pointing to correct "cluster_id"

# ambari-server stop
# ambari-server reset
# ambari-server start

.

[#5]

@Jay Kumar SenSharma

IT Worked pefectly !!!!

A heartily thank you for your valuable support.

Appreciate the way you tackle the issues.

Appreciate your way of providing perfect solution with smooth writing skill.

[#6]

@Rohit Sharma

Thank you 🙂