during upgrading from HDP 2.6.2 to HDP 3.1, we faced some issues "starting 2 data nodes was giving OOM exception, so we had to increase the java heap size manually and start them manually as well" that required manual steps , now we are at "Finalize Upgrade Pre-Check", and everything looks fine in the cluster, so we need to bypass this step in the upgrade, as its giving us error below
Upgrade did not succeed on 2 hosts Your options: Pause Upgrade, delete the unhealthy hosts and return to the Upgrade Wizard to Proceed. Perform a Downgrade, which will revert all hosts to the previous stack version.
The above was originally posted in the Community Help track. On Sat May 11 02:59 UTC 2019, the HCC moderation staff moved it to the Cloud & Operations track. The Community Help track is appropriate for questions about using the HCC site itself.
Hi @Tarek Abouzeid,
Not sure if you where able to bypass this issue ,
But I have few troubleshooting checks that might help
Check 1 : we should navigate to problamatic two nodes and see if the hdp-select command lists the version 3.1 or not.
Check 2 : navigate to ambari-database , inspect the table :
select * from host_version order by host_id; select host_id,host_name from hosts;
and you will know two of your culprit hosts will be in OUT_OF_SYNC state .
make it current based on the repo_version_id and restart ambari.
Hope this helps
I'm seeing the same issue. One of the hosts is in a "installed" state as opposed to "current". How exactly do I go about "make it current based on the repo_version_id and restart ambari"?