What is our primary use case?
I work with Telcos, one of the cell phone providers in South Africa, and they use it for their billing infrastructure.
The applications running on the WebLogic suite are for billing and customer CRM, which goes out to the call centers and the dealers. We maintain the environment.
What is most valuable?
The feature that I have found to be the most valuable is the ease of deployment.
In the beginning, you do tend to struggle a bit, but once it's deployed, then everything falls into place and maintaining it is quite simple.
What needs improvement?
It is difficult to say which features can be improved at the moment, as we are not working with the most current version and I am not aware of the features offered in the new version.
Once we catch up and move on to version 12C, we need to see what can be migrated to the cloud.
It might not be suitable to migrate the systems to the cloud, or maybe only portions of it. For example, it makes sense for our web services to go on the cloud, but not the actual application, the CRM system.
If we are considering the version that we are currently working with, then I would say that it's all fairly straightforward when it comes to using it. However, there are some small things, such as being able to restart clusters, where you can choose to restart each server one by one instead of all at the same time.
The ripple start is what we refer to as shutting down and restarting one server at a time in a cluster. In other words, when you kick off a ripple start, and it would go through, it will shut down the one instance, and start it up, then it would move to the next one. It wouldn't shut them all down, and I wouldn't have to manually, stop one, start it up, wait for it to come up and then move down to the next one. This solution would benefit from the inclusion of a ripple start function for clusters.
Also, the cloud integration, which I've heard is very strong with Oracle, it's the shift and lift methodology.
IBM WebSphere used to do things like that, where you could do a ripple start as opposed to shutting everything down and it would manage each one individually. That would be useful. if it's a live environment we have to ripple start. That's the big one, otherwise, we are pretty happy with everything.
The debugging function is nice on the Weblogic, but one thing WebSphere has, is, that you can apply the debugging permanently, or just until the server is restarted.
That might also be a feature that would be nice on WebLogic, but not critical because we turned it off afterward.
For how long have I used the solution?
I have been using this solution for ten years.
What do I think about the stability of the solution?
This solution is very stable.
The only time it's not stable is when the code has a memory leak, or it's heap dumping or the garbage collection isn't fine-tuned. That is not the environment, it's the code. The environment itself is extremely stable.
We have to get caught up as the version we are using is out of support.
The buzzword right now is cloud, and at some point, we have to see what we can take to the cloud and what we cannot. There are plans to move in that direction.
What do I think about the scalability of the solution?
It is scalable. We have added extra servers and extra instances when it's been required.
We don't run on VMs, we run on IBM LPARS. We don't run VMs where you can have them firing up, on-demand, but it is scalable for our purposes.
How are customer service and technical support?
Officially it's not supported, but we do get support when it's required. For example, approximately six months ago there was that day-zero vulnerability bug that had to be patched.
The patch that we applied on WebLogic actually broke some environments.
We logged tickets and worked with Oracle and they were able to support us, isolate the issue, and give us new fixes.
The support was very good and worked very well.
From this experience, I would rate the technical support quite highly. They were able to pinpoint the issue quite rapidly and assist us with a new patch. I would rate them a nine out of ten.
Which solution did I use previously and why did I switch?
Previously, we used the IBM product called WebSphere.
WebSphere and WebLogic are both very similar. They have the same purpose, the same end. I liked the way WebLogic is compartmentalized in the server where you can go and find the configurations, and see it on a file. It's fairly file-based, the data source is everything.
WebSphere wasn't stored quite that way, so you couldn't work as nicely outside the system.
There might have been a few other tweaks that WebSphere had which Oracle doesn't. But on the whole, I would say Oracle is far better, it more superior to the IBM product.
How was the initial setup?
The initial setup is complex. We did a migration from WebSphere to WebLogic.
The reason it's complex is it was already running, but it was a very different animal than WebSphere. There were code changes required, which fell to the developers on the development side. On the operational side, things like fine-tuning little things like the data sources work a bit differently, but once you figure one out, then the rest all falls into place.
At the moment the deployment model we use is on-premises, and nothing has been migrated to the cloud. It's a project for the future.
The deployment was approximately just over one year to get it migrated fully to where we were stable enough to turn off the WebSphere.
It was a little bit better than I had expected it to be. We all felt it would be an eighteen-month to a two-year project, and it did come in a little bit less than that. But of course, the business expects it in three to six months. We did try but realized that it was not going to happen unless everything just magically works the first time.
I'm on the operations side, I'm not on the development side. We look after the infrastructure and the upgrading.
The developers are a large team. On the operations team, we have approximately ten people. One person can do a feature release, which is what we call a deployment, in an evening. This is done three times a week.
What about the implementation team?
We do deployments roughly once a week, three times a month.
We have our own in-house developed deployment manager, which we call the Deployamater, and they set up all the deployments. The manager fetches the EARs, JARs, pages, and JSP files, then it deploys them.
We don't use the automated deploying via Oracle. We manage it like that, but we do it in an offline environment.
We duplicate our environments and we go to our offline environment, deploy there, test it first, and then switch the traffic to the new environment that it's being deployed to.
What other advice do I have?
I am a subcontractor to Vodacom, and the company I work for is a vendor, and they are an approved vendor with Oracle.
It is difficult to offer advice because every scenario is different, but I would suggest that you use the available expertise. There is a lot of expertise, don't try to do it all alone.
I wouldn't go back to WebSphere and for me, I would say it is the market leader.
I would recommend this solution and I would rate this product a ten out of ten.