Overall Satisfaction with GraalVM
We are in an investigative stage with GraalVM and AOT (native-compile) at the department level (software engineering and SecDevOps). We have deployed test applications written in Java and Scala to Windows, Ubuntu, and as Docker images running on container services (Google Cloud Run and Kubernetes).
At this point, I am very impressed. There was a learning curve to understand GraalVM, the polyglot features, and native-compiles. Pluralsight has a good fundamentals overview. The home site for GraalVM has a number of good "getting started" guides.
The areas that we are seeing major impacts in are software size, load time, and cold start time. These are critical benefits as we try to minimize the size of our containers and reduce cold-start time. The end result is more container services can be deployed with the same hardware. This could have a significant financial benefit over running code in a JVM.
We are also investigating the Polyglot features, such as running Node.js and Python in GraalVM. The ability to combine Java and Node.js could present some interesting features to combine the best of our libraries.
In my opinion, this product and technology will significantly improve the performance and footprint of Java-based containers to serverless containers, Kubernetes, and OpenShift.
At this point, I am very impressed. There was a learning curve to understand GraalVM, the polyglot features, and native-compiles. Pluralsight has a good fundamentals overview. The home site for GraalVM has a number of good "getting started" guides.
The areas that we are seeing major impacts in are software size, load time, and cold start time. These are critical benefits as we try to minimize the size of our containers and reduce cold-start time. The end result is more container services can be deployed with the same hardware. This could have a significant financial benefit over running code in a JVM.
We are also investigating the Polyglot features, such as running Node.js and Python in GraalVM. The ability to combine Java and Node.js could present some interesting features to combine the best of our libraries.
In my opinion, this product and technology will significantly improve the performance and footprint of Java-based containers to serverless containers, Kubernetes, and OpenShift.
- Reducing software cold start time
- AOT--Native compile, which removes the dependency on the JRE and JVM
- Reduces application size. Combining native-compile and EXE compression significantly reduces the size of our containers
- Native compiles. We have problems figuring out how to package jars with all dependencies for compilation. This might be our problem as we progress through the learning curve.
- More support and testing for Sprint and GraalVM.
- I recommend providing installer packages instead of downloading zip files.
- Financial: Increased Java container count on the same hardware.
- Performance: Decreased cold start time for container startup.
- Training: Adopting GraalVM has a learning curve. This requires investment in time and resources. The benefits come through re-evaluating our current deployments for optimization.
Do you think GraalVM delivers good value for the price?
Yes
Are you happy with GraalVM's feature set?
Yes
Did GraalVM live up to sales and marketing promises?
I wasn't involved with the selection/purchase process
Did implementation of GraalVM go as expected?
Yes
Would you buy GraalVM again?
Yes