So far in the series we have looked at how the integration platform was envisioned, how it evolved through projects and a recap of where it was vs where we thought it would go. The next step is to consider how the platform might evolve in the future. This means considering some of the technologies which might play into what we do.
BizTalk Services vNext
BizTalk Services vNext is a big area we are keeping an eye on. While in the platform so far we felt that the value offered by the features in current BizTalk Services did not overlay well with the requirements we had and also the bits where it might, the cost of BizTalk Services couldnt justify us using it until this point, BizTalk Services vNext is something we have high hopes for.
The main thing around BizTalk Services vNext that we are interested in are as follows:
- Can we reduce our investment in BizTalk Server and move some of it into BizTalk Services
- Can we use BizTalk Services as the back end to some of our API projects
We expect that BizTalk Services vNext will not offer a complete alternative to BizTalk Server. This would be fine. We would be happy if BizTalk Services vNext let us focus on those solutions where BizTalk Server would typically struggle or where we would need to have a very large setup of BizTalk Server. An example of this could be where we have chosen to only use a small setup of BizTalk Standard edition so far. If our requirements brought in the need for BizTalk Enterprise for the full high availability and higher throughput scenarios, we would hope that BizTalk Services vNext workflow offering could be something we would consider as an alternative. In these kind of requirements the operational costs to support a more powerful and critical set up of BizTalk Server would be something we would like to off load parts to Microsoft where we would not need to worry about 90% of the things that can be setup in efficiently now with the Server version.
We are quite bullish about the opportunities we hope BizTalk vNext will offer and we would look to potentially migrate some of BizTalk Server processes to services.
API requirements have been quite a large set of use cases for us across our projects. Microsoft is investing quite a lot in enhancing its API offerings. This is one of the areas we will keep a close eye on so that we can look to take advantage of those new features to help us build better API solutions.
We have already made a reasonable investment in the usage of Azure IaaS. We will need to track the changes Microsoft introduces in this space to support our continued effective use of this. We could sit on our laurels and just continue to use what we do now, but that is not really in line with the principles of how we work. We know that as Microsoft introduces new features we might be able to improve our platform by either taking advantage of features which let us do new things or taking advantage or features which may reduce our operating costs.
Perhaps the biggest thing we will be looking for in this space is around the support for fail over clustering on Windows in Azure IaaS which would improve our BizTalk Server story. If you like to learn more about Azure IaaS you might want to check out somewhere like https://www.thousandeyes.com/solutions/azure-monitoring for more information.
Business Intelligence is one of those areas within our current platform we would like to improve. Because we have a decoupled architecture it is difficult to get a holistic view of how process execution crosses systems. We need a business activity style monitoring approach. We feel however that BizTalk Server Business Activity Monitoring is not a good thing to invest too much in for us within this platform at the present time. If you are an integration shop using just BizTalk and not much else then BizTalk BAM is a decent choice but in our platform BizTalk is just one part of the platform and while we could still use BAM in BizTalk Server and move events from other areas into BAM we would hit some constraints:
- The BAM Portal is ugly and makes BizTalk look old
- The BAM data model isnt really going to scale to internet volume
- BAM in BizTalk has a painful ALM experience
Based on these three factors we have chosen not to use BAM so far but instead would be looking to take advantage of the wider BI capabilities on offer elsewhere in the Microsoft stack. On my old blog I wrote an article about how we might do this, and technology has probably moved on again since then too. We will be keeping an eye on this space and technologies such as Event Hubs and HDInsight and Power BI type technologies which could give us a great way to leverage insights into our processes.
Azure Active Directory
At present we are using Azure AD a little bit to support the integration platform, but we are also using it elsewhere in the organisation to underpin our Dynamics CRM investment. Azure AD seems to be having a lot of investment from Microsoft at present and it is something we feel that will have more and more information coming into the community domain about how to best take advantage or it. This will be something we will be keeping a close eye on.
One of the biggest drivers within our platform is to reduce the operating costs. We already have automation of some areas of the integration platform but we will also be monitoring this space to keep up with technology changes which can help us to increase the level of automation we can do. The knock on effect of this should be to help reduce our operating costs.
Currently we do not have any requirements in the pipeline for IoT style projects. That said this is something we will keep an eye on because there is lots of innovation happening in this space. Down the line this area could be something where opportunities for innovation exist in our industry and it would be good for our platform to have an awareness of whats possible so that we can be ahead of the game if this comes up.
As architects this is one of those topics which falls into the area where it might not be part of your business strategy now but you need to 2nd guess some of what you think might come down the line so you can provide a service which allows integration to be an enabler for great solutions. I expect in due course we will probably invest in some proof of concepts in this area.
Hopefully this will give you an insight into some of the thinking we have about how we will be keeping an eye on the technology space to see what things could add value to our integration platform. We are big believers in continuous improvement and are always looking for ways to improve the way we work or the things that we do so that we can take advantage of any opportunities that might present themselves.
Please remember that the above thoughts are relevant as of this point in time. The rate of change in the cloud space is high and you can only see so far ahead into the future. There could be new things which could come up later which would be better for our integration platform or likewise Microsoft could reduce investment in a particular area which means it is less likely to be something we can use in the short or medium term. To deal with this we need to ensure we keep an awareness of whats happening in the industry but also build solutions which are adaptable and easy to change.