What is most valuable?
Here are the top three features:
- Deployment workflow (WF) can be designed this way, so that it is not necessary to provide all applications (systems) artifacts of which an application consists. Presence of deployed artifacts may differ in every deployment and an Automic ARA tool is able to handle this situation and provide a final successful installation. When we were testing different ARA solutions for this use case, it was necessary to design more than one deployment WF (specific for every combination of artifact). Our goal is to design a suitable amount of workflows for every application. In an ideal case, just one complex WF will solve all requested deployments. Until now, we have been able to keep this principle, that we have only one WF for each application.
- Ability to easily create our own Action Packs for platforms where we are deploying artifacts (a wide range of supported platforms out-of-the-box). You can read and edit out-of-the-box action packs as well. The tool is very good, configurable and customizable for specific requests of customer (specific customer environment and processes).
- Well and clearly designed UI with a very good UX. User is able to very clearly see designed WF, WF progress during deployment, and output logs from deployment targets and platforms. All this is available inside the Automic tool. Work in the tool is intuitive. User is able to restart deployment after it was stopped because of some problem, and so on.
How has it helped my organization?
CA Automic ARA solution gave us the ability to:
- Be faster and more flexible. In first pilot run, we achieved five times faster deployment without human work.
- Be effective. We are saving tens of millions of dollars of work from our IT administrators.
- Support standards and have fewer outages. Environments have to be the same because we use one deployment workflow for each integrated application.
- Support Agile development. We are able to do quick sprints without any deployment limits.
Our DevOps project (reason why we bought this solution) has improved culture and cooperation in our IT department significantly.
What needs improvement?
GUI for mobile phones: Availability to approve and start deployment through mobile phones.
Better dashboards and reporting: About deployed versions, planned deployments, waiting deployments, and so on.
Infrastructure discovery functionality: Scan customer infrastructure automatically and collect information about potential deployment targets (purpose is not to insert all the information manually). For example, XebiaLabs has this feature.
For how long have I used the solution?
We have used this solution in our production environment with our first application since April 2017. Actually, we have eight applications integrated in this solution.
What do I think about the stability of the solution?
Solution is stable, without any platform outages. We have had a few bugs, but technical support solved those quickly.
What do I think about the scalability of the solution?
No, we did not have any issues with scalability. Our solution runs in HA mode, with two active nodes. It is enough for our usage and current load from integrated applications.
How are customer service and technical support?
Bugs are very quickly fixed by the vendor. Until now, it has only been a positive experience from our point of view. Also, a platform upgrade was solved with an onsite vendor consultant. Activity was difficult, but it was done without any problems.
As a general statement, we can say that we are satisfied with CA Automic support in all areas that we are cooperating.
Which solution did I use previously and why did I switch?
We have used only custom deployment scripts (each team has their scripts, we did not have any standards for that), for a solution like this. Major reason for this purchase was our DevOps project. As a short introduction to this project, we started our DevOps transformation at our bank and we were solving problems like build automation, provisioning automation, test automation, and deployment automation as well. We wanted to have one complex DevOps delivery pipeline which consists of all of these interconnected tools, and we made it.
How was the initial setup?
Initial preparation and installation took three months and was very complex. We wanted to prepare the HA and a stable environment, which would need to be ready for massive usage. Actually, we have already integrated eight applications and run over 1000 automated deployments.
What's my experience with pricing, setup cost, and licensing?
They should use a deployment target (or agent) licensing model and start their implementations with a lower number of agents in the beginning.
Which other solutions did I evaluate?
Yes, for sure. We did more PoCs during the analysis phase of our DevOps project last year. We were testing products: IBM UrbanCode, CA Release Automation, XL Deploy from XebiaLabs, HPE Codar and the winner - Automic ARA.
What other advice do I have?
This product is very strong and, on the other hand, very complex. So they should involve more technical people in the beginning to ensure knowledge transfer, and continuous and quick integrations of applications into this solution. The best option is to create a dedicated automation team before the implementation.
The second important thing is that this solution provides only automated deployment and some orchestration as well, but deployment depends on all steps before, such as build and artifact preparation, testing, and provisioning. So others should use at least the automated build and artifact preparation and interconnect it with deployment. Then, they could achieve bigger benefits from automation.