Executing a Successful Migration to v7.1.2 During Our Busiest Time of Year

[Skunkworks69]

• Executing a Successful Migration to v7.1.2 During Our Busiest Time of Year
• Why did we upgrade?
• What was our planning process?
• What steps did we take to complete our migration?
• No code changes were needed
• What was the most challenging aspect of the upgrade and how did you address it?
• How are we doing now?
• Embarking on an upgrade? Here are a few tips:

Executing a Successful Migration to v7.1.2 During Our Busiest Time of Year

By Gary Folland 
Director, robotic process automation, 
Australian government organization

A large Australian organization serving 26 million citizens has been using RPA for more than four years. Gary Folland manages the RPA capability, and his team runs the digital workers. Before the organization upgraded to SS&C Blue Prism v7.1.2, they looked to speak with other customers about their experience. In that spirit, he wants to share his upgrade experience and a few tips and tricks with the broader SS&C Blue Prism Community.

Why did we upgrade?

We were using Blue Prism version 6.8 and decided to upgrade to v7.1.2. Our primary reason for upgrading was the enhanced security and support offered by the new version. Getting access to additional features that were available with 7.1.2 through Hub was also part of our technology roadmap. We’re aiming to extend the scope of automation use cases in our business. Some use cases couldn't be fully automated because we couldn't put a human in the loop, and we planned to solve that challenge with the extended capability of Blue Prism Interact.

 

What was our planning process?

We spent time planning, understanding our options for the upgrade and identifying the pros and cons of each method. Blue Prism had a detailed upgrade pack that highlighted the different pathways we might take to complete our upgrade.

We determined the best path for our organization was a parallel migration. We planned to set up a parallel infrastructure so that we could run both versions of Blue Prism at the same time. Once we had temporary services and virtual machines, we could set them up separately to make sure the new version was working properly in our environment. We would migrate our current automations across to the new environment, and once we were certain everything ran properly, we’d decommission the old environment.
We participated in a valuable two-day workshop with our IT group and the Blue Prism professional services team to iron out the details about how to best accomplish the migration.
We had a project plan that detailed the security approval, the logical access model, architecture documents, and how to conduct the review with our central audit logging team — and then how to get those items reviewed, endorsed and approved. This project plan drove what we did and ensured that we were comfortable as we reached each new stage.

As a government agency, we don't like to be on the bleeding edge of anything, so we wanted to speak with another customer who had successfully upgraded. The professional services team put us in touch with a customer who could vouch for a successful implementation; being able to confirm that knocked down our last barrier to organizational support for the upgrade.

What steps did we take to complete our migration?

Timing was tricky for us. Much of our work is seasonal, and some processes only occur once a year. We didn’t necessarily want to conduct an upgrade at our busiest time of year, but we had to test certain automations during that busy period because they don’t run at other times. We ended up waiting until the initial flurry of activity had died down and then we began migrating processes.

Once we were ready to go, we:

• Spun up the new infrastructure and replicated our application and database services.
• Spun up an additional 30 virtual machines, which are our bot runtime resources.
• Reviewed the performance of the machines running 7.1.2 and made sure the solution was working as we expected.

Once we were confident everything worked, we started to move the automations off the old infrastructure and put them into the new infrastructure so that we could have a rollback if we needed it. If things didn't work as expected, it would take us three seconds — the time it takes to push a button — to return the automations to version 6.8. This way, we didn’t lose any capacity and could continue to run our automations. At the start, we had 101 machines running version 6.8 and 30 running v 7.1.2. As we filled up the 30, we started to convert some of the 101 that were left into 7.1.2 machines once we had some space.

No code changes were needed

We didn’t change the target applications that the code base is running on. Only the Blue Prism infrastructure and application changed. We found that the code worked just fine.
We started off thinking maybe we’d move one automation, and then give it a couple of days to see if it was working. But everything worked so well — and it was so easy — that we began putting two or three automations into the new infrastructure each day, quickly testing them to confirm all worked well. We initially thought it would take us five or six weeks to complete the migration, but we did it in just three weeks.
During the installation phase, I had help from Deloitte along with our own IT people. Whenever our IT group was unable to progress, I’d reach out to my contact at Deloitte and for assistance to sort out any issues and make sure that the team was able to keep making progress. In turn, our Deloitte contact reached out to his colleagues in Blue Prism professional services to vet and validate decisions. This was critical for us; our IT team didn’t have much installation experience since Deloitte had managed the initial installation and the first upgrade.

What was the most challenging aspect of the upgrade and how did you address it?

We had some challenges with the configuration settings. Our IT team downloaded the installation guide from Blue Prism and just followed it through, but when things weren't working, they didn't know why. We called our Deloitte and Blue Prism experts, who helped us tweak the configuration settings. Next time around I’ll plan to have a detailed conversation about the installation guide with Blue Prism and ask specifically about the configuration settings.

How are we doing now?

We have 42 processes live in v7.1.2 and I just did a shout out to my team for their thoughts on how 7.1.2 is doing…here is what they said (unfiltered):

• “ooh, Friend!”
• “7.1.2 is better QOL for a lot of things, new features and a bit faster running things”
• “it can defs handle control room and sessions better as well”
• “Resources are available immediately after logging out, helps with scheduling too”
• “it used to take like 5-10 mins to expand the schedule list - now it shows up when we open bp”

Happy team … happy head of RPA! We took the opportunity to re-stack our automations during the migration, have optimized utilization rates and capacity, and our first Interact automation has moved into phase three of Hypercare which are all very positive outcomes for us.

Embarking on an upgrade? Here are a few tips:

• Take your time planning your upgrade. We benefited from internal conversations with business stakeholders and executives, from engaging with other Blue Prism customers — and from the workshop we held with the Blue Prism professional services team.
• Consider a parallel migration. This helps with contingency planning and allows you to roll back any processes that aren’t working the way you expect. This made it a ‘no-brainer’ for our exec groups to understand how no automations would be interrupted.
• Don’t be afraid to upgrade at a busy time for your business. We did this and were able to test processes that are only live at that time of year. We knew that if we ran into trouble, we could call our partner, professional services or simply roll the processes back to the original infrastructure.