With the public arrival of the Cloud Software Group (CSG) and news of how it will operate in future the announcement by one of its key constituent parts, TIBCO, of its latest steps down the product roadmap gain an extract frisson of importance. This comes from the fact that CSG is now the holding company for both TIBCO and Citrix, but also for NetScaler, Spotfire, IBI, Jaspersoft, XenServer and ShareFile as independent constituent business units, with each having its own identity to continue its own line of innovation.
Given that Spotfire, IBI and Jaspersoft have long been seen as integral parts of TIBCO it will be interesting to see how the company's roadmap will alter and develop. None are expected to shoot off tangentially, but all might feel some pressure from the Group’s core large enterprise marketplace for future tweaks and additions that were not part of original roadmaps. It is possible to see a bit of that thinking appearing in the latest developments from TIBCO, announced at the company's recent conference in Munich.
The key one, around which the others all make their contribution, is the introduction of a higher-level, single-pane management and control environment, which is meeting the demand that has come for such a move from customers.
It is certainly a move that makes good sense for the company, as the TIBCO's General Manager, Ali Ahmed, observed during his conference keynote:
When enterprise customers make investments they are looking not three years ahead, but anything up to 20 to 30 years ahead. That puts a different kind of planning pressure on the company that not many other vendors face. If we go down, Wall Street goes down, FedEx will not be able to deliver packages. You know, airlines like Lufthansa, Air France KLM. United Delta, get grounded.
In addition, those enterprise customers are looking at business re-invention over the long haul. This will need management of all types of data, in forms that may not yet be known or fully understood. This could be metadata reference data, master data, or transactional data; therefore providing a single unified platform for managing this and accommodating the changes as they come along, on-premise or in the cloud, will be a major requirement.
Given that long haul future, and the organizational changes it and Citrix have undergone, it is also possible to look at it as move to provide stronger, richer levels of management as three of TIBCO’s key components get a new degree of liberty and freedom of action.
Removing pain with a pane
The goal here is to provide a central management available capable of running on any cloud service as well as on-premise. Providing a single place where users can monitor and manage, It will have all the capabilities of messaging, integration with BusinessWorks, event processing, data virtualization and data management with the EBX Master Data Management system. As well as providing day-to-day management, it will also provide management services for strong developer tooling, to oversee the adding new services or taking remedial action should problems occur.
All of this will be managed through a single 'pane of glass' as the control plane operating across multiple cloud services – be they private clouds hosted locally or as co-lo operations, or resources located in public cloud services, and legacy on-premise systems. It will work seamlessly across multiple operating systems, going back to the many different flavours of Unix. The aim is to give users control over all that they utilise, wherever it resides physically, just how `legacy’ it is, as Ahmed pointed out:
The only thing that I am saying here is to come and meet you where you are in your cloud. It's not our cloud. It's not the TIBCO cloud anymore. It's going to be your cloud.
He claimed that TIBCO has been the only vendor offering identical functionality across all platforms, because it has always shared the codebase. It is now taking this same philosophy to the cloud, whether it is AWS, Azure, GCP, OpenShift, or any of the growing number of vendor-branded cloud services popping up. The aim is to be as platform-agnostic as possible, a factor which is likely to be more important to its current customer base than any new customers. Current customers are obliged to work with a mix of legacies as well as new cloud native systems, he noted:
The theme here is radical simplification, we want to really simplify the packaging and the pricing involved and make it easier for you to do business with us and reduce the number of available SKUs around our core capabilities. And you minimise the cross-pricing complexity.
Rajeev Kozhikkattuthodi, TIBCO's Product VP, and Matt Ellis, who heads the platform product management, teamed up at the conference to lead the delegates through some of advantages that fall into developers’ laps from using the single pane management environment. They will be able to take metrics and correlate traces from across all services, and correlate the results to the logs and the metrics associated with a service. So you can quickly solve problems, get your systems back up and running if something were to happen, and also do root cause analysis much easier than possible before, said the duo:
We're building a control plane that will enable you to view all of the observability metrics, as well as `stoplight-style’ monitoring, in a single interface that spans across your entire infrastructure, whether your infrastructure is Kubernetes, running on-premise, whatever it is. Wherever it's running, all of it is available in a single user interface. We're giving operators the governance to lock down what version of BusinessWorks, or BusinessWorks Container can be deployed, and what version of plugins can be used. This should guarantee working with the smallest container image possible that Hawk works with known, tested versions.
This exploits TIBCO’s existing distributed monitoring and management software, Hawk, to build what Ellis referred to as `The Control Tower’, a single binary with no dependencies required, that drops into an existing infrastructure. From here, it can automatically discover and connect to all running applications, and then begin securely sending this data to the controller, via a tunnel to the control plane.
As for applications running on-premises, the pair indicated that a simple general rule applies when it comes to ensuring they work correctly in such a potentially complex environment. It is, essentially, `do nothing’. They do not need to be refactored, they do not need to be migrated, and they do not need to be lifted and shifted.
Down to the edge of the edge
The single pane approach also plays to the company’s existing presence in edge computing. According to Kozhikkattuthodi, TIBCO has had most of the component parts needed to provide edge computing services as parts of other applications and tools for several years, but has never tried to bring them together as a coherent whole. The management capabilities now available to it come together with the introduction, some seven years ago, of Flogo.
This is TIBCO's fully supported implementation of the open-source Project Flogo Framework, and perhaps THE key component of its edge toolset. It is designed to allow developers to build ultralight, event-driven micro-services, and uses browser-based flow design technology to allow applications to be compiled down to web assembly level, allowing users to start running logic and browsers on any device that has a suitable processor chip on-board. This means that business process logic and system control functions can be distributed down to the very end point devices of the network.
The new TIBCO management platform will allow all the elements of Edge computing to become an integral component of enterprise environments. The company sees this providing the potential for detailed, in-depth forensics, in real-time and at the point of occurrence down at the edge, in the same way as currently possible with forensic applications running on major data centers. With the addition of AI and ML services at the edge this could soon include fault analysis and implementation of remedial actions from data centre to the furthest corners of the network edge. The same capabilities can also be applied to the provision of effective governance services right out to the network extremities.
A strong response to large enterprise customers’ lobbying. TIBCO's 'single pane of glass’ management environment not only covers the operations of large and complex environments on a soup-to-nuts basis of the largest data center to the furthest corner of the corporate network, but also the integration of the workloads of developers as they introduce upgrades and additions, or tackle and remediate problems with service delivery