With the hype trains, the exaggerated promises, and the inevitable under-delivery that seems to accompany most trending tech these days, it can be quite tough to work out what’s legitimately important and where things actually are in terms of market adoption.
Such is the case with microservices, the trending architecture approach where systems are broken up into distinctly available and independent services. Carriers and vendors seem in broad agreement that it matters, as evidenced by the level of blog and hashtag activity. But thus far, established vendors don’t have a lot to offer in this area.
Why exactly is that? I think it is at least partly because microservices architecture really does represent the kind of tech disruption that doesn’t come along too often, and vendors are still processing what exactly it means for them and for the market.
It also takes investment at significant levels. More than a facelift of APIs to expose Web services at finer levels of granularity, it is an architectural style that often involves refactoring and rewriting across the entire application stack. It means embracing a different philosophy around coding, packaging, releasing, and deploying. And it requires a different kind of organization to support it, with a different mindset and culture.
That’s why microservices adoption is an area where newer vendors can have a real advantage over the incumbents (whether they can exploit it to unsettle the status quo of the market remains to be seen.) It’s also an area where we see carriers lead the way.
This doesn’t mean that the microservices approach can’t be applied to older systems; it certainly can. But like those incumbent vendors that tried to package their client/server screens into ActiveX to “Web-enable” them in the face of the startup Web-native competition, trying to bolt on microservices may do more harm than good in the long term. Especially if the refit is superficial and the goal is hashtag enablement and RFP box-checking.
Given this, it’s probably unreasonable to expect core vendors to be very far along in terms of adoption. And that’s the reality: They aren’t, despite what the blog posts, tweets, and hashtags might sometimes suggest.
That’s not necessarily a bad thing either. Microservices architecture isn’t the right solution to every problem, and there are still open questions about whether it has the broad applicability in core administration to justify the hype. Insurers should understand what value microservices do (and do not) bring to their infrastructure, how they would utilize a microservices-enabled system, and where that need fits into their prioritizations when evaluating solutions.
Add new comment