Upgrade your Python version
If you are on ThoughtSpot-managed CentOS hosted on AWS, Azure, or GCP, your ThoughtSpot cluster comes with Python 3.9 by default. You do not need to upgrade your Python version; ThoughtSpot does not support any Python version other than Python 3.9 for ThoughtSpot-managed CentOS clusters hosted on AWS, Azure, or GCP that are on 8.4.1.sw.
If you are on customer-managed RHEL or customer-managed OEL, or ThoughtSpot-managed CentOS hosted on VMware, SMC, or Dell, Python 3.6 is still the default version. If you are on customer-managed Amazon Linux 2, Python 3.7 is still the default version. If you are on customer-managed RHEL or Amazon Linux 2, you can upgrade to Python 3.9.
ThoughtSpot supports use of the following versions of Python 3 for deployment. Python 3.7 is the current default version for customer-managed Amazon Linux 2. Python 3.9 is the default version for ThoughtSpot-managed CentOS hosted on AWS, Azure, and GCP. Python 3.6 is the default version for customer-managed RHEL, customer-managed OEL, and ThoughtSpot-managed CentOS hosted on VMware, SMC, and Dell.
Python 3.6: ThoughtSpot-managed CentOS for AWS, GCP, VMware, Azure, SMC, and Dell; customer-managed RedHat Enterprise Linux (RHEL); customer-managed Oracle Enterprise Linux (OEL); customer-managed Amazon Linux 2
Python 3.7: customer-managed Amazon Linux 2
After you create the cluster, upgrade your Python version to Python 3.9.
To upgrade your Python version, follow these steps:
sshinto a node as an admin user.
Run the following command to stop your cluster:
tscli cluster stop
Ensure that the version of Python you want to use is ThoughtSpot-certified for your deployment type. ThoughtSpot does NOT certify Python 3.9 for customer-managed OEL or ThoughtSpot-managed CentOS for VMware, SMC, or Dell.
Install Python version 3.9.
Navigate to the correct location to run the command:
Run the following command on all nodes, at
/usr/local/scaligent/release/for each node:
PATH=/usr/pgsql-11/bin:$PATH pip3 install -r production/prod_image/current/pip_packages/requirements_no_versions.txt --upgrade
Run the following command to start your cluster again:
tscli cluster start