BQ1 - What is the Cloud Transformation Program?
A program of work to prepare UNSW for the future of compute. Providing a common vision, direction, and approach for consuming cloud services and enable UNSW to transform and accelerate digital service delivery. Phase 1 encompasses the exit of the government data centre (GovDC) and migration of existing workloads to Cloud Services (AWS).
BQ2 - What is the Business Readiness Assessment (BRA)?
The BRA is a series of questions that allows the migration factory to understand the current workload and state of servers running in GovDC. With the assistance of UNSW business and technical owners, it established server workload baseline and enables the migration path determination.
BQ3 - What comes next after the Business Readiness Assessment?
The information will be used to formulate a wave plan, a group of servers or workloads, to be migrated to Cloud Services. The completion of this assessment phase marks the readiness to commence the T-80 process (see migration overview)
BQ4 - What is the T-80 process?
T-80 a reference to the process applied to the migration of workloads to cloud services, typically conducted over eighty (80) days. It encapsulates the associated elements for the migration, including application discovery, environmental readiness, user access, dry runs, technical validation, formal testing, approvals, cut-over, and decommission.
BQ5 - How will the AWS hosting costs be handled?
Existing GovDC workloads will be migrated and hosted in AWS free of charge. Over time, as customer requirements change and UNSW start to leverage the true benefit of cloud services, customers can build to their own requirement and budgets. If you never change your current environment after migration, the service will remain at no charge.
BQ6 - Who will be taking care of the testing leading up to the migration cut-over?
The Cloud Transformation Program (CTP) have on hand Testing Analysts from within UNSW and our partners. They will work with application business and technical owners, to validate and test their environment prior to the T-0 cut-over. This is done around T-30-T-20. Business owners and technical owners have the option to include external testers (e.g. vendor or alternate team) to perform testing.
BQ7 - Will my application work in the Cloud?
Yes. The requirements and handling is clearly established during the BRA and T-80 discovery stages.
BQ8 - What will happen if I don't do anything?
UNSW's existing data centre, GovDC, is due to be retired and all existing servers need to be migrated from GovDC onto the AWS Cloud. If the server isn't migrated before the retirement of the GovDC data centre, the applications running on the server will cease to function.
BQ9 - Will AWS training be provided before the migration?
There is no planned AWS training currently. The Cloud Platform Engineering team will be managing the hosts and underlying infrastructure, but feel free to reach out with any questions.
BQ10 - I understand the default migration pattern is to rehost, what is involved in a rehost?
A rehost is a 'lift and shift' of the existing server(s) onto AWS. This involves a continuous, block-level data replication of the server onto AWS into a sandbox environment. Before the cut-over, we power-off the server in GovDC, then move the server in AWS out of the sandbox onto the production environment and complete a DNS cut-over.
BQ11 - I understand some servers are being replatformed, what is involved in a replatform?
A replatform involves rebuilding the server on AWS. This requires significantly more effort. The Migration team can provision a new Red Hat Enterprise Linux 8 host or a Windows 2019 Server host, but the technical owner/vendor will need to reinstall all the required applications. It is generally reserved for servers running on unsupported versions of Linux or Windows that need to move onto AWS.
TQ1 - I currently take snapshots of the server on VMware before any major updates. Will I be able to do the same with AWS?
We don’t yet have a process for system admins to go into the AWS Console to backup/restore from a backup at the moment (as we are currently leveraging automatic daily full backups on AWS, which run at 08:00 AM UTC), however we are still building out the processes as end users continue to define requirements. At the moment if you need to manually take a snapshot or restore from a backup the Cloud Platform Engineering team at UNSW will be able to help with that.
Please raise a ticket with the IT Service Desk and assign it to the IT.Cloud Systems Support group.
TQ2 - Will I have the same MAC address, IP address, and domain name(s) on AWS?
The MAC address and IP address of the server on AWS will be different to the server on GovDC, however it will retain the same domain name as we complete the DNS cut-over. It is important to ensure that any MAC address/IP address references to the server and database on GovDC be updated.
TQ3- Will there be any down-time during the migration to AWS?
The requested outage window is twelve (12) hours, however the typical outage for a replatform is only three (3) hours and a rehost is four (4) hours, with a provision to roll back should there be any unexpected issues.
At 9:00am we turn monitoring off, then at 9:15am we shutdown the on-prem host and migrate. If everything goes well by 2:00pm the application is migrated, if there are challenges the migration could continue to 5:00pm and could be followed by a rollback. Migrations do not need to start at 9:00am if the business requests an evening migration or weekend migration, but they introduce overtime.