The key’s in the API, not the ERP, for today’s organizations to develop
End-users are an indiscriminate bunch. If the service they’re trying to access or have subscribed to is slow, or worse, slow and unreliable, then it’s the app or service that gets the bad reviews. In many cases, the app is robust, solidly built, well served, and is the answer to every users’ prayers— it’s the access to the service that’s causing problems.
That detail, of course, is of no interest to end-users. It’s not significant to a paying subscriber if API calls take an age, are under-resourced, and haven’t been stress-tested properly before full production. The details simply don’t matter, but poor API management can taint an entire brand. It seems like a significant oversight — after all, no organization would consider placing a domestic ADSL router-come-firewall-come-Wi-Fi-access-point in front of a production web cluster. But that’s the mentality that’s limiting API management in today’s increasingly microservices-led environments.
RIP the ERP
The business world has changed significantly in terms of the type of technology running organizations and the services they provide. Large, behemothic software packages have lost their dominant position in the control of business processes and data and are being moved aside in many enterprises to a more compartmentalized role, or a handful of roles. Specialized, smaller, more agile apps are being dropped into place in discrete business functions to achieve specific tasks, and these are being interconnected via their APIs, and — to varying extents — business processes are being automated by orchestrating data movements between apps and data silos.
Development operations, too, are working similarly. Discrete containers and microservices are creating environments where application developers can (almost) use building-block development methods. Microservices and automated deployment apps like K8s let DevOps teams work in cycles of continuous development, providing shifting business strategies with agile, quick-to-production services.
Arise, Sir API
Adhering to the “ancient” UNIX credo of doing one simple thing and doing it well, interconnecting highly-capable apps and services at scale is the task of the API, but exposing services with a “fire and forget” mentality is a relinquishing of responsibility on the developer’s part that simply doesn’t wash. DevOps functions are not discrete and separate from the rest of the enterprise, and the organization’s needs in its strategic goals (serving the end-users, for example) means that every department must think of the bigger picture. That’s why significant thought must be given to API development, deployment, management, monitoring, and iterative improvement — like any other element of what IT does.
Like the brand being tainted by the poor overall performance of a key app or service, the professionals writing amazing, lean, and powerful code (and those professionals’ managers) will be significantly let down by only considering APIs as an afterthought. It’s time the API got the attention it deserves.
Here at TechHQ, we’re looking at three platforms that help companies provide the type of APIs that their offerings deserve: lightning speed of transaction, low latency, agility, scalability, reliability, and security. Primarily, we’re interested in API platforms that provide the type of framework of support that’s devised with the DevOps professional in mind. That means vendors’ teams that speak the language of both the developer and the business-focused stakeholder, too. Issues of scalability and security must be addressed by API platform vendors as well as questions about OAuth and plugins! We think the following three companies are ahead of the game in both regards.
Tyk started as a community development project and its engineers remain the heart and soul of the company, to this day. Tyk’s API Gateway remains a free and open source product, so users are permitted — indeed, encouraged — to trial, configure, use, and deploy the platform before they buy.
The company monetizes by offering cloud-native and multi-data centre versions of the Platform, whether self-managed or Saas. But the support the company offers is its real USP, which is entirely done by the company’s product engineers. That means that whether you’re a start-up launching the next big thing in IoT or a huge enterprise with a sharding or cluster issue for mission-critical API traffic, you get access to the same level of insider expertise
This understanding of the importance of a helping hand coupled with a hyper-focus on developer experience means that the Tyk API Gateway and Management Platform is highly configurable to individual companies’ uses at every stage. From creation and testing of APIs to deployment in production, the Tyk platform helps companies expose their APIs and services in a controlled and reliable manner. The platform scales to enterprise levels, and combines a business-focus with the protocols, languages and technologies that DevOps professionals are familiar with.
(*It’s pronounced “Tyke,” by the way, to rhyme with “Mike.”)
The Anypoint API Manager from Mulesoft is a largely Java-based API development and management system that sits happily, alongside the rest of Mulesoft offerings. Anypoint is enterprise-level API management, so expect a price tag to match, although there are three tiers of prices for the more cost-conscious.
The platform helps build APIs and distribute them multiple times, grouping services and apps into a single dashboard, wherever those services and apps may be: cloud, hybrid, or in-house. Anypoint is available as an IPaaS, but however it’s deployed, the framework comes with templated security and policy templates, so you hit the ground running.
APIs are assigned a type in terms of deployment (production, development, and so on), so entire business functions’ API use can be managed, monitored, and configured easily.
The abstraction capabilities of the MuleSoft solutions mean that APIs to microservices’ or behemothic apps’ APIs are all handled in the same way. Users can proxy any API in order to manage (or monitor) traffic better, and alerts can be configured either to respond to security issues or according to load balancing thresholds.
APIs can be created, rolled out, and even provided with documentation via templates, thus encouraging better service use and safer services for all. Click here to learn more about this top-grade offering.
The AMPLIFY API Management platform helps teams create and publish APIs from cloud or on-premise server instances, and then provides management frameworks to monitor and control all APIs across the enterprise.
There are built-in analysis tools in AMPLIFY so that users can see live operational data, or even developer uptake metrics (API “popularity contests,” if you will).
The analysis routines can flag anomalies and report on overall API stack health, or users can use “traditional” logfile analysis, Splunk, and so on — the AMPLIFY API Management system is as extensible as an API!
Axway is a heavily business-focused outfit, and the AMPLIFY platform can be used to monetize API use easily, making APIs reusable: develop once, deploy (and charge) multiple times. That facility is via AMPLIFY Central Service, which centralizes monetary issues like charges and subscriptions of every and any API across the enterprise.
Axway is building a complete developer and user community for its products, with an internal marketplace available to all subscribers where code and methodologies can be shared, thus vastly improving time to market for new services’ APIs.
To learn more about Axway, click here.
*Some of the companies featured are commercial partners of TechHQ