Alright motherf*ckers,
After reverting back from snapshot, gave it another go, was following exact same procedure from the link above: “Manual HDP Stack Upgrade: HDP 2.2.0 to 2.2.4″
This time, I’ve decided to start with step 8, before step 7 under “Performing a Manual Upgrade” – Restarting clients using API right before restarting all other services.
The curl code is a bit wrong, therefore, make sure you first copy the entire command to notepad and fix ‘{ “RequestInfo”: … }’ with ‘{ “RequestInfo”: … }’ <— NOTE THE TICKS!
Also, the whole curl command line is missing “-d” switch right after the URL, and should look like this:
curl -X POST -u username:password -H ‘X-Requested-By:ambari’ http://ambari.server:8080/api/v1/clusters/MyCluster/requests -d …
When I started restarting services, first failure was HIVE with the below error:
Fail: Execution of ‘export HIVE_CONF_DIR=/etc/hive/conf.server ; /usr/hdp/current/hive-metastore/bin/schematool -initSchema -dbType mysql -userName hive -passWord [PROTECTED]’ returned 1. 15/05/15 17:23:28 WARN conf.HiveConf: HiveConf of name hive.server2.enable.impersonation does not exist
Metastore connection URL: jdbc:mysql://edgenode01/hive?createDatabaseIfNotExist=true
Metastore Connection Driver : com.mysql.jdbc.Driver
Metastore connection User: hive
org.apache.hadoop.hive.metastore.HiveMetaException: Failed to get schema version.
*** schemaTool failed ***
2015-05-15 17:23:29,871 – Command: /usr/bin/hdp-select status hive-metastore > /tmp/tmpYexwiz
Output: hive-metastore – 2.2.4.2-2
To fix this, I’ve started mysql server manually and re-run the ‘Restart All’ command for the HIVE which now completed successfully!
Later I failed to start RANGER with the below error:
Fail: Execution of ‘/usr/bin/ranger-admin-stop’ returned 126. -bash: /usr/bin/ranger-admin-stop: Permission denied
To fix this one, I used ‘chown -R ranger:ranger /usr/hdp/2.2.4.2-2/ranger-admin’ to fix the permissions and re-run the ‘Restart All’ command for RANGER which also completed successfully!
When I was done with all services and restarts, executing ‘ambari-server set-current –cluster-name=MyCluster –version-display-name=HDP-2.2.4.2′ failed.
Now all I had to do is to restart all ambari-agents on all hosts and following by restarting the ambari-server.
Waited few minutes for all agents to register with ambari-server and this time I was successfully able to update the cluster version!
Hope my two-cents help,
autogun