Target audience: Mid to Highly skilled Linux, IBM and GCP admins.
Currently IBM software is retrieved from passport advantage client account. If you’re tasked with building out multiple IBM products on GCP VPCs and Compute Engine instances then setting up an install media repository that is fast and local is a must.
This article describes how to uses your GCP bucket as an IBM install software repository. Once the IBM install software repository is available then the Compute Engine startup section can be used to dynamically customize each Compute Engine instances. Use the compute engine startup to install IBM software prerequisites, and then install and configure the IBM software.
The process goes like this:
- Retrieve the clients IBM software from passport advantage.
- Upload the IBM software(install media) to the GCP bucket in the Region(s) where the IBM products will be installed.
- Upload install and configuration scripts for the IBM software to the GCP bucket.
- Use FUSE to mount the GCP bucket.
- Setup your GCP Storage policy so these large install media files will transition through the GCP storage object life-cycle. Basically move them from GCP nearline to coldline. The objective is to reduce cost of storing these large install files and still keep copies for auditing and creation of a DR site. Once the repository is setup, now create a shell script that.
The Shell script will:
- Installs and additional packages(RPM files) to meet the IBM software pre-reqs.
- Use the software repository mount point to retrieve the install media to the compute engine instance.
- Unpack install media files.
- Run install via response files.
- Configure the IBM software.
- Test you setup.
References:
