Platform Core: Azul Zulu Builds of OpenJDK
Improved cold startup insights and JVM warmup analysis have highlighted heavy infra needs
What is our primary use case?
My main use case for Azul Zulu involves running our platform on Kubernetes, where all our applications are written in Spring Boot and Java, and we face the problem of cold startup. The pods come up very late while traffic starts coming in, so to solve this, I interacted with the Azul team, who suggested using it to address the cold startup and JVM warmup problem. I want to add that we focused on the warm-up aspect.
What is most valuable?
One of the best features Azul Zulu offers is its ability to reduce compile time for byte code loading by almost 40 to 50% during our testing.
The ease of integration was quick with Azul Zulu, as they provided custom images of the JDK which we pointed to in our clusters, and the setup was really quick.
We did see faster JVM startup times, but the overall latency in JMeter showed that our API response times were not improving.
Positively, I decided not to go with the solution because for our single-tenant application, the entire cluster we spun was for 40 vCPUs or 100 gig memory, while Azul itself was demanding almost around 40 vCPUs for a 40 to 50 vCPU cluster, which meant doubling the infrastructure, resulting in heavy costs, which is why we dropped it.
What needs improvement?
Based on my experience, the major area for improvement with Azul Zulu is the infrastructure requirements.
I think having more knowledgeable support personnel involved in the exercise with companies would be great instead of having salespeople pitch in.
For how long have I used the solution?
Recently, we did a POC for Azul Zulu because our Java application was having a cold startup problem, and we wanted to solve that issue.
What do I think about the stability of the solution?
Azul Zulu is stable based on my experience during the POC.
What do I think about the scalability of the solution?
I do not see any scalability issues with Azul Zulu; however, I noticed that as the size of the cluster increases, the infrastructure requirements from Azul also rise, meaning if I'm running a two vCPU pod, it almost requires a two vCPU allocation for Azul Zulu as well.
How are customer service and support?
I had multiple interactions with their support team, and they were able to help us effectively. I would rate the customer support an eight on a scale of 1 to 10.
Which solution did I use previously and why did I switch?
Previously, we used CPU Q-Boost, which could dynamically double the CPU to handle cold warm-up problems during startup, allowing us to solve issues we faced with multiple spikes, and being an open-source solution, it met our needs.
How was the initial setup?
Azul Zulu is deployed on-premises in our organization, as they provided us the images, which were hosted on some cloud, and we deployed them in our cluster after getting necessary security exceptions from our SOC team.
What was our ROI?
I have not seen a return on investment; however, I want to highlight that there was a developer who carried out the entire exercise independently, and the ease of use was good, which potentially saved time and employee resources.
What's my experience with pricing, setup cost, and licensing?
The pricing was a concern, and ultimately, we had to drop Azul Zulu due to pricing and infrastructure concerns.
Which other solutions did I evaluate?
Before choosing Azul Zulu, we evaluated Q-Boost and Azul Zulu, but ultimately, Q-Boost won the race.
What other advice do I have?
On a scale of 1 to 10, I rate Azul Zulu a six.
I gave it a six because, in my situation, for my company and software, it is not able to fully support our needs, and the claims made about the product were very high while the actual delivery was underwhelming, as I felt the salespeople were too aggressive in their commitments.
During my POC, I found the reliability and accuracy of Azul Zulu's output to be good, especially regarding JVM internals where there was improvement.
My advice for others looking into using Azul Zulu is that if you have a good budget and are a big company with heavy clusters that can afford the price, Azul Zulu is a very good option that can solve cold startup problems, particularly in JVMs.
Rapid automation has accelerated backend development and has transformed project delivery speed
What is our primary use case?
My main use case for Azul Zulu in the current project is developing the archiving and unarchiving system. I use Azul Zulu on the back end when I created a REST API, which I use together with Java Spring Boot and the libraries. I do everything through the IDE, so I use it to build, together with OpenAPI, my REST API.
How has it helped my organization?
Azul Zulu has had a positive impact on my organization because I am new to the company and they asked me to develop with recent technologies. They use pure Java, without Spring, and when I brought the solution, they liked it a lot, mainly because of the speed at which the project started to be delivered.
I have noticed measurable results after adopting Azul Zulu. I joined this project which had already been underway for approximately six months, and in two months, I refactored the whole project and wrote it, and it is functional. During the six months that the project was in the hands of another developer and with another technology, the project practically did not move. Then with the adoption, in two months, the project moved forward significantly.
What is most valuable?
In my opinion, the best features that Azul Zulu offers are speed, as today we need to develop applications very quickly, and the speed that it allows us to implement is what helps me.
Regarding speed, it is not specifically one feature but the whole set of features that benefits us. Sometimes I have worked on other projects with Java that took time, and some things are not automatic. Processes that are automated and files that are generated by the IDE or by the CLI itself help a lot in development instead of having to build thousands of lines of code that, theoretically, are generated by a command. It already comes ready.
What needs improvement?
I think Azul Zulu can be improved by always thinking about productivity, as most languages tend to evolve in that direction. I have two projects that I am working on in parallel, one using Angular 21 and the other using Angular 22. If I use CLI commands in Angular 21, it will build the files with the extension; for example, it will create dot component. If I use it in Angular 22, it no longer creates that. Developers noticed that this was a performance gain in terms of visualization. Everyone already knows that is a component. The same applies to Azul Zulu. Developers have to observe the development community a lot and have an open channel for improvements so they can enhance the way developers create, not just thinking about the final product. The final product is extremely important, obviously, but for construction, developers have to be prepared for that technology and understand that it is easier for them, that it speeds up the work. If you simply put something in, and that feature you added is not applicable, the developer will not use it, making it unnecessary work.
For how long have I used the solution?
I have been using Azul Zulu for about two years.
What do I think about the stability of the solution?
Azul Zulu is stable in my experience so far.
What do I think about the scalability of the solution?
The scalability of Azul Zulu in my experience seems easy and fast, though that belongs to another department.
How are customer service and support?
Azul Zulu's customer support is not handled by me.
Which solution did I use previously and why did I switch?
I used a different solution before Azul Zulu, and I changed because it was requested. As a developer, they requested that a specific stack be used, and that is what I did.
How was the initial setup?
My experience with Azul Zulu's pricing, setup costs, and licensing is that all of that was handled internally. It was not done with me, as I am just the developer and do not have access.
What was our ROI?
I cannot present any return on investment metrics with Azul Zulu because it is not my department. The company I work for is a public agency, so that is not available to me either.
Which other solutions did I evaluate?
I did not evaluate other options before choosing Azul Zulu.
What other advice do I have?
My advice to others who are thinking about using Azul Zulu is to try to explore as much as possible what it offers, getting away from the conventional paradigm. I would rate this review as a nine out of ten.
Standardized our Java estate and have reduced licensing costs while maintaining strong support
What is our primary use case?
Azul Zulu facilitates cross-platform compatibility as we use it as the base for our Docker images in our CI/CD pipeline. For instance, we migrated our legacy Spring Boot applications from Oracle JDK, and Azul Zulu served as a seamless drop-in replacement that allowed us to maintain the same performance level without any code changes.
What is most valuable?
Azul Zulu positively impacts our organization by providing a platform that allows us to standardize Java on a well-supported platform. Without the high cost associated with other proprietary vendors, Azul Zulu provides an excellent platform for running our applications.
What needs improvement?
For how long have I used the solution?
What do I think about the stability of the solution?
What do I think about the scalability of the solution?
How are customer service and support?
How would you rate customer service and support?
Positive
Which solution did I use previously and why did I switch?
What was our ROI?
Since Azul Zulu is open source, the number of licenses needed has dramatically decreased, which has further contributed to our return on investment and led to continued productivity gains.
Which other solutions did I evaluate?
What other advice do I have?
Which deployment model are you using for this solution?
If public cloud, private cloud, or hybrid cloud, which cloud provider do you use?
EC2 instance
Good one to start with TF journey .Happy to review it and start with the cloud journey .Lets deploy the 1st instance using TF