Step 3 of a Mammoth Upgrade Fails With Curl Errors When Uninstalling Big Data SQL

(Doc ID 2315932.1)

Last updated on OCTOBER 14, 2017

Applies to:

Big Data Appliance Integrated Software - Version 4.6.0 and later
Linux x86-64

Symptoms

Step 3 of a Mammoth upgrade (in the example here 4.6 to 4.9) fails with curl errors when uninstalling Big Data SQL. The failure looks like:

Step 3: Shutting down Hadoop Services
...
INFO: Uninstalling Old Big Data SQL. This will take some time ...
...
ERROR: Puppet agent run on node bdanode03 had errors. List of errors follows

*************************
Error [14488]: (//bdanode03.example.com//Stage[main]/Bigdatasql::Deletebds/Exec[deletebds]/returns) change from notrun to 0 failed: /opt/oracle/BDAMammoth/bdaconfig/tmp/deletebds.sh &> /opt/oracle/BDAMammoth/bdaconfig/tmp/deletebds.out returned 1 instead of one of [0]
*************************

1. Drilling down through the /opt/oracle/BDAMammoth/bdaconfig/tmp/deletebds.out shows that the failure is due to a failing curl command on bdanode03.example.com. The failure looks like:

# /usr/bin/curl -L -u admin:<admin_pw> http://bdanode03.example.com:7180/api/v1/cm/commands
curl: (35) SSL connect error

2. Comparing the curl version on the host where the command fails with a host where it succeeds shows a difference in the NSS version like below. On the node where curl fails NSS is 3.21 on node where curl succeeds NSS is at 3.27.1.

a) On bdanode03 where the curl command fails: 

 

Cause

Sign In with your My Oracle Support account

Don't have a My Oracle Support account? Click to get started

My Oracle Support provides customers with access to over a
Million Knowledge Articles and hundreds of Community platforms