What is our primary use case?
Currently, we are using UrbanCode Deploy, a product from IBM. Before I had a bit of expertise with it at a bank in Belgium that uses recent commissions from computer systems.
Now, I work in Belgium for an importer of Volkswagen, Audi, Porsche, Lamborghini, etc. I'm in the middleware department. What we do is we link all the different applications together.
We deploy UrbanCode as a middleware solution.
How has it helped my organization?
In the beginning, this tool was completely new to us, to everybody in the company, actually. At that time, Urban CodeDeploy was not performing well at all. After that, when we started to use the tool more and had some better insights, UrbanCode Deploy started performing better.
Last week we sat together with some IBM experts to see where we could gain some more deployment type of performance optimization. We still have to do this. We got new recommendations yesterday. We will now see how much effort it will take to implement them. We will not implement them if it takes over 2 weeks.
If we can gain maybe ten or fifteen percent optimization with UrbanCode Deploy, then we will invest further in the platform. The tool is performing not so bad now.
What is most valuable?
For me, the most valuable thing is that we have a full overview of our environment, where we have deployed what, when it went okay, when it went wrong, etc. These kinds of things.
It is very easy to make a software release. It used to take us at least a couple of hours to make a release, now we went to production with a new one last night. This new release took me five minutes.
It makes it easy to select what we want to deploy and to follow it up afterward. The reports that we get when something goes wrong work very well.
What needs improvement?
What UrbanCode Deploy improved for us is that we have better visibility. We are not deploying faster with it. Actually, it's a little bit slower.
We have the software application automation fully under control with UrbanCode Deploy. Whereas when we did it the old way, we had failures.
Before we used UrbanCode Deploy it was difficult to find out why we had failures. Now, we don't have any failures anymore, or almost none.
For the last couple of months, we have no failures at all and we have full visibility. The interface is doing very well.
Performance-wise, our deployment now takes longer. We know that, but that's something that we can live with to take advantage of UrbanCode Deploy's other feature set.
We are one major version behind and we will do the updates internally. The only problem is UrbanCode Deploy is used not only by me but by all of the company. We all do our deployments with UrbanCode. Upgrading is not my decision.
I certainly would like to have a better way to pass information between deployment steps using UrbanCode Deploy because that's really difficult to do.
Personally, I would like it if you can loop on some things like the tags and the agents. I would like to have some more loops.
For how long have I used the solution?
One to three years.
What do I think about the stability of the solution?
I'm really not that satisfied with the stability of UrbanCode Deploy. First of all, it needs a lot of memory and it needs a lot of CPUs. This has a huge impact on viewing. I expected an interface that is far more responsive and it seems most of the screen handling is done on the server and not on the client. Hence, just for working with it, you need a machine with 4 CPU's or cores to be comfortable.
If you use the API's to do some stuff from Jenkins it performs way better. Doing the same action using the REST API, is literally just hundreds of milliseconds, whereas the GUI takes seconds. That is certainly something that UrbanCode Deploy could improve on.
What do I think about the scalability of the solution?
The current version that we have, the 6 series, does not scale at all. It is not good enough, in my opinion. They knew this and they have upgraded this with a whole new technology. We do not need it to scale any further at this moment.
How are customer service and technical support?
In the beginning, we had some issues with UrbanCode technical support, some things that did not work the way we thought they would work. For some of them, we opened PMR's with IBM. As always with IBM PMR's take a long time to respond.
I'm not happy with UrbanCode customer support at all. On the other hand, we also put some time in with these experts last week. These experts were certainly highly qualified. With them, we proceeded very well and very quickly on some items, so it's a little bit mixed.
The support that we get from PMR's is not good. I give it a score of 2 out of 10. We have PMR's open for more than a year.
Which solution did I use previously and why did I switch?
Management wants to use UrbanCode to make sure that we all deploy the same way. I do not know if they evaluated other products. I was just given the product at the moment and I was in the company already for some time. The strategy behind this is that everybody deploys with UrbanCode.
What about the implementation team?
We have a department called Infra. Management decided to use UrbanCode for several reasons so it was imposed on us that we should use it and it was installed outside our team.
Which other solutions did I evaluate?
Because the company likes to use the big solution providers, our middleware is IBM. Our deploy tool is IBM so they have a lot of big players. I personally prefer Open Source solutions. I don't just want to use Open Source. We use Jenkins & GitLab.
Jenkins & GitLab are tools that everybody uses and if there are problems, these problems are fixed quickly by the community or by the developers. That's basically how Open Source works. I am not at all against paying for a solution. For example, it could well be that we use GitLab but the Enterprise Edition. It gives us the opportunity to start with a community edition.
I like to evaluate tools to see how they work, to see how they fit us as a company, and then switch to a paid solution. This is something I miss with big companies like IBM.
What other advice do I have?
In my opinion, the way you have to develop with UrbanCode Deploy is to use the graphical interface, which is nice. However, for development UrbanCode Deploy is slow.
Each time you have to click on something, you have dropdown menus, etc. I am a coder by nature. Some tasks I would normally prefer to go faster. It would be better if I could type commands directly using UrbanCode Deploy.
I would give UrbanCode Deploy a six or a seven because it is not easy to set up. The documentation is not commonly found. I think it's pretty expensive too. UrbanCode Deploy is not so much used in the market. When you implement, it's not so easy but you can overcome that. Afterward, when it runs, it runs slowly to complete. Now that we have the solution, we rely on it and really trust it now. I would have given it more points if it were easier to work with. My final rating would be seven out of ten.
Take at least one month and properly evaluate the tool with a real business case because it was sold to us in a different way. We were told this tool can do anything and within a week you will have a solution. In reality, it took us three or four months even to come to the final solution. The tool is not as easy as the vendor says it is. You have to get used to it and you should not think that this tool is just magic that will solve your problem. You have to think around your current problem and then see how you can get that into the tool and not vice versa. Don't expect the tool to solve your problems.