I'm part of a significant communication company in Europe, and we primarily use IBM App Connect to integrate our CRM system.

One of the most beneficial features is the ability to handle multiple communication technologies, like integrating Kafka flows, which is helpful as other teams heavily use it. Regarding error handling, I initially wrote most of the handling myself. While built-in features for error handling are built-in, it largely depends on the developer. We use a custom solution that catches all exceptions, logs them in a database, and replays them as needed. It has been effective for us for over twenty years.

I've been using IBM App Connect for about twenty-five years, and while I like it, there are some areas for improvement. The trace policy is ridiculous, and the biggest issue is the cost—it's expensive. People would use it much more if the price point weren't so high.

I've used IBM App Connect since around 1999 or 2000 when the solution was called MQSI. I've been using it for approximately twenty-five years now.

App Connect has been incredibly stable over the twenty-five years we've used it. Any incidents were usually related to other components like databases rather than the App Connect broker.

I haven't explored scalability much as we operate on a stable platform and haven't delved into containers or cloud setups yet.

I found IBM's tech support surprisingly good. They responded promptly and delivered a patch within a week for the most recent issue we faced.

Positive

Regarding similar products, occasionally, we've had proposals for other products, but I find IBM App Connect to be the most compatible and reliable for our needs. When comparing it to Kafka, for example, I prefer IBM App Connect because of its ease of deployment, control, and stability.

For IBM App Connect, deployment and installation are straightforward for us. We use Jenkins scripts for automated functionality, making the process easy. It typically takes around forty-five minutes to complete all the steps, including dropping the existing broker version, installing the code, setting up links, restarting with new code, creating a new broker, and deploying.

One engineer is enough to handle the deployment process, as it's automated with Jenkins. Maintenance is minimal, and upgrades are usually straightforward. We encountered a bug with the MQTT model in version twelve, but IBM provided a patch and a script to test it, and the team completed the upgrade in fifteen minutes. Overall, maintenance and upgrades are relatively easy with App Connect.

As for new features, I haven't explored all of them yet, but I appreciate the plug-and-play nature of App Connect with different interfaces.

Overall, I would rate IBM App Connect as nine out of ten, but with the disclaimer that I haven't tested all features extensively. The usual explanation for any drawbacks relates to the pricing and justification for the cost.

We use it to transform data, route data, and validate data. So, App Connect is an enterprise service bus (ESB), like Mule, Fuse, or Appian. This means it's really an enterprise service bus, and we use it for integration. We interconnect applications and protocols with App Connect.

I like the adapters. The adapters help us achieve scalability. If you want to connect to SAP, there's an adapter. Salesforce? There's an adapter. You want to connect to another system? There's likely an adapter for that. 

You want to connect to various services like Gmail, email, FTP, SFTP, MQ, JMS – anything you want, it's likely included in App Connect through the adapters.

The main advantage is that it's a central point. If we compare it to a country, it's the post office.

There is room for improvement in the logging messages. Sometimes, if you put someone new into App Connect, they can abandon it the same day. The logging is really painful. 

However,  IBM has made efforts to integrate with Elasticsearch for logging, so that's an improvement. Overall, the logging can be difficult at times.

One more important point is that if IBM improves its CI/CD capabilities, it will make a big difference. Right now, I have to create my own CI/CD setup from scratch for every client, which is inefficient.

Back in 2013, I worked with Sonic ESB, and even then, it had CI/CD with Maven. With App Connect, you need to build everything yourself when using tools like Jenkins, Bamboo, or CircleCI. IBM really needs to provide official support for this.

I have been using it for nine years. 

I've had stability problems in the past, but the issues were usually due to configuration errors.

When we're able to prove that the problem is with the tool itself, IBM sends what it calls a 'fix pack' to address the specific issue. 

We never had scalability issues. We always install it in cluster mode, and it's been very reliable. I've never had a problem scaling it up.

I would rate the scalability a nine out of ten. 

It's the typical IBM support experience. Sometimes, you send a message, and you can tell it'll be handled in India or some other company. 

You get the feeling that some support staff are just waiting to go home, but sometimes you get a really good person who follows through. It's inconsistent.

Support is definitely a weak point.

Neutral

Every company I've been to, I had to redo everything. I connect to their Linux system – just dev, not even quality, UAT, or production. They give me dev access, I connect, and I test things.

it's not that the setup is difficult, but there are many factors. 

I send an email explaining the issues I find. Twice in my life, this email went all the way to the board with the CIO and others. They questioned my findings, and after I provided detailed explanations, they asked me to fix the problems.

There are two main things. First, App Connect is very strict software. You need to read every line of the documentation. If the documentation says 'don't do this', then don't do it. People sometimes don't have time and skip around in the documentation.

Then, I come in and find issues. Like, I've been to a company – a big insurance company managing millions – where there were no passwords or restrictions on their production App Connect and MQ systems. I asked what was going on!

This is just one example. I went to another company where they changed the file access rights ('chmod' in Linux) on MQ, even though the MQ documentation specifically says not to touch them or that it will break. It's written in the PDF, and someone changed it anyway to make their life easier. But those changes create problems because you're paid to manage complexity, not avoid it.

So, installation is not easy, and there are many options. Also, you need to understand the contract you signed with IBM. If you signed up for active-active clusters, don't set it up as active-passive. IBM can take legal action if you violate the contract terms. And so on.

It's the same for MQ because MQ is strongly linked with App Connect. They can be installed independently now (that wasn't possible in old versions like IIB), but many clients only wanted the ESB. So, IBM separated them. There's a strong relationship between MQ and App Connect, and some App Connect features require MQ.

There are two different versions of App Connect Cloud – Smart Connect and one that's similar to the on-premises version. So, it can be confusing for someone new to it.

For a good ROI, you need to find someone who works well. If the person isn't skilled, your ROI will suffer. The problem is there aren't many of us with this expertise.

If a company has a good individual or, better yet, a good team, then App Connect works well. But many young people don't want to put in the effort. It requires a wide knowledge base - Salesforce, SAP, SFTP, HTTP, security, and more. They'd rather specialize in Java or .NET.

With any ESB (not just App Connect), you need patience and a willingness to learn. It's a common problem with other ESBs like Fuse and Mule, too.

App Connect is not cheap. But there might be other products like Cisco that are even more expensive.

My recommendation depends on lots of factors. Things like architecture, design, existing IBM products, and budget all play a role in the decision. For a small company, I might suggest something cheaper.

But App Connect is a very good tool. 

Overall, I would rate the solution a seven out of ten.

Our company has a big project for delivery and e-commerce sites. We use IBM App Connect to deliver information to our end users after they buy our products. We get information from many places, such as enterprise solutions and other applications. We send information and gather information from them to use in our web applications. We give our end users updates on the status of the products they bought from us.

The solution is very fast-paced when it comes to making things right. We can quickly download what we need if we need any app migrations for our software business. We can download it plenty of times when we need to use it. It gives us very high speed for delivering any development. 

The tool's automation capabilities have given us great automation abilities, especially on the factory side. We have many integrations, creating value by using this capability. We're canceling many manual processes and migrating them to IBM App Connect automation. This is improving and helping our way of working and developing many integrations.

We can handle many complex integrations that were difficult with other solutions. In other tools, we often needed to do a lot of coding and write many things from scratch. But with the solution, many things are already ready in the platform. This gives us high confidence and saves time, especially for complex integrations.

For simple tasks, the difference between products might not be obvious. However, IBM App Connect makes it easy in complex environments where we take data from many different applications, send it to another app, and apply algorithms. It's useful for these more complicated integrations. The solution is also more mature than other products. 

Finding developers for the product is difficult since it is a niche solution. I know the OpenShift environment is running well for the microservice environment. We had some issues with the other environment we tried to implement. It can be easily implemented internally, but we have some problems in practice.

I have been working with the product for three years. 

The solution is generally stable from an application perspective. If regular maintenance, such as patching and upgrading to the latest enterprise versions, is done on time, you should be on the safe side with minimal issues. However, neglecting updates on the web server could lead to problems.

The tool is scalable. However, there are considerations regarding tool licensing and costs. Companies worldwide are focusing on cost reductions, which can impact scalability. While the current licensing might be sufficient for existing applications and APIs, adding new APIs and developments could increase usage statistics, necessitating additional licenses. This would mean extra costs, which might bring about different considerations from management. My company has 100 users. 

If you have experience with other similar systems, the tool's deployment is straightforward; I can say that. But if you're new to it, it can be difficult to understand which menu you must download, what kind of files you need, and many other things. It's not easy to understand at first. However, after one or two installations, it becomes very straightforward to do further implementations and installations.

Our deployment strategy for IBM App Connect is part of our larger plan to move all applications to the cloud. However, we face some challenges, especially in Europe due to GDPR and in Turkey due to local laws like KVKK. These regulations make moving to global cloud providers like AWS, Azure, or Google Cloud difficult.

Deploying to some local providers isn't easy, and this is a major consideration in our strategy for installing IBM App Connect or any other applications in the cloud. We must consider this and arrange all other steps for solutions accordingly.

Deployment is usually easier with global cloud providers like AWS, Google Cloud, and Azure because their service management works well with App Connect. As I mentioned before, App Connect Enterprise supports OpenShift very well, which is powerful.

However, in some cases, we need to expand the environment for our applications. This is another point to consider when planning our cloud deployment strategies. The deployment process can be completed in two weeks, depending on the environment we're working with. If we're working with a local cloud provider, it can get more complex. This is because local cloud providers often don't have experts familiar with OpenShift, so we need more parties involved in installing IBM App Connect in the cloud.

However, if we work with a global cloud provider like AWS, we can install with a smaller team. Maintaining the tool isn't difficult if it runs on a stable infrastructure. The maintenance process is straightforward in a secure and stable environment. However, maintenance becomes more complex if the infrastructure is unstable or insecure.

It's worth to work with the product in the long term. 

IBM App Connect's pricing is high compared to other products. 

If you're not missing any business opportunities, IBM App Connect is a great tool to work with as an enterprise service for managing many tasks. It is excellent for developing APIs, integrating various services, and simplifying development processes. It offers easy development and integration capabilities, making it a highly recommended tool. I rate the overall solution a nine out of ten.