Cisco Acquires AppDynamics – Implications
On January 24, 2017 Cisco announced its intent to acquire AppDynamics a leading vendor in the Application Performance Management market. The price was announced as approximately $3.7B in cash and assumed equity awards. This acquisition is likely to shake up the management software business at all layers of the stack.
A Brief History of Application Performance Management
The APM industry was born along with the invention of the Java language and the first Java run-time environments known then as J2EE Applications Servers. Wily Technology was founded by Lew Cirne (remember that name) in 1998.
Wily Technology was acquired by CA in 2006, and was at the time the market leader in the APM industry. In the same approximate period of time, IBM, HP and BMC all acquired companies in the APM business, making APM into a feature of their enterprise management suites.
Dynatrace was founded in Linz Austria in 2006, and was the first APM product to focus not just upon the performance of the Java applications in the JVM, but to focus upon entire end-to-end transactions with its PurePath technology. Dynatrace was acquired by Compuware in 2011. Thoma Bravo (the private equity firm) acquired Compuware in 2014, and then spun Dynatrace out into its own (still privately held company). Dynatrace has been listed as a leader in the APM market by Gartner for the last couple of years. Dynatrace also gets credit for starting the “modern” APM market by being the first of a set of innovators that remade the market, and kicked IBM, BMC, HP and CA out of the Gartner leaders quadrant.
Lew Cirne founded New Relic in 2008 just as public clouds were getting going. New Relic brought several innovations to the APM space. New Relic was the first APM product delivered as a service (SAAS). It was the first vendor to support a language other than Java (Ruby). And it was the first APM vendor to ink a distribution deal with a public cloud (want fries with that big meal). New Relic went public in 2014. Gartner currently lists New Relic as a leader in the APM Magic Quadrant.
Jyoti Bansal, the former VP of Development for Wily founded AppDynamics in 2008. The crucial innovation of AppDynamics was a way to trace end-to-end transactions across the tiers of an application system in a way that worked out of the box for most modern web applications, and in a way that imposed minimal overhead out of the box (with no configuration required). AppDynamics has been listed by Gartner as being in the leaders quadrant for APM for the last couple of years.
The Recent History of the Systems Management Business
The success of Dynatrace, AppDynamics and New Relic has effectively turned the APM products of the major systems management vendors (IBM, BMC, HP and CA) into legacy offerings. The three new vendors have been capturing the lion’s share of the new application deployments both on-premise and in the cloud and have been eating into the installed bases of the legacy vendors as well.
The problems for the big four have however not been limited to the APM space tamiflu online. According to Gartner, all four vendors have been suffering from declines in market share in the IT Operations Management space as a whole as shown by the Gartner chart below.
So the story in the systems management business is that the Business Service Management strategies of IBM, BMC, HP and CA have all failed and that monitoring in particular has devolved into a best of breed business. See the graphic below for our attempt at depicting the current diversity of the monitoring business.
The Best of Breed Monitoring Industry
The Modern Full Stack Monitoring Problem
All of these changes need to be analyzed in light of how the monitoring problem has changed over time, and continues to change at an accelerating pace. Applications are now changed frequently in production (due to Agile and DevOps). Many different languages and runtimes are in use. And everything (compute, networking and storage) is subject to multiple levels of virtualization that create abstraction, opaqueness and management challenges.
The image above gives rise the question, “Exactly what approach stands any chance of covering applications from end-to-end (from the inception of the transaction to its completion), across the entire stack (from the click of the mouse to the reads and writes on the storage and back) while keeping up with the pace of innovation, the diversity of the environment, the layers of abstraction and the increasing levels of automation”?
Most people think of Cisco as a vendor of networking gear (switches and routers). But this is not the whole story especially as it pertains to Cisco’s acquisition of AppDynamics. Cisco announced its Unified Computing System in 2009 and has since then become a market leading vendor of servers and converged infrastructure to the enterprise. So when one thinks of the impact that Cisco will have once it owns and integrates AppDynamics into its product portfolio it is essential that one look at Cisco as the primary competitor to IBM, Dell/EMC and HP Enterprise in the business of supplying compute infrastructure to large enterprises worldwide.
It is clear from the Gartner chart above that BMC, CA, IBM and HP have not done well in management software in the last few years. Dell has the advantage of owning VMware whose vRealize line of products has been doing well of late, but the Quest assets (acquired by Dell) are not state of the art. For example, Gartner does not even list Dell as being present in the APM market in its latest Magic Quadrant.
So with the acquisition of AppDynamics, Cisco has created a huge problem for two sets of vendors. One set is its competitors in the enterprise computing and networking business (IBM, Dell, HPE) who now lack by comparison any kind of a compelling end-to-end performance story. The second set is the legacy management vendors (IBM, BMC, HP and CA) who now have to contend with an AppDynamics funded by Cisco, and assisted by Cisco’s formidable sales force.
The challenge that Cisco and AppDynamics will face is that any attempt to redirect the R&D efforts of AppDynamics towards tighter integration with the UCS and Cisco switches will inevitably result in AppDynamics spending less time maintaining its differentiation in the APM market. For an example of what could happen look at what happened to BlueStripe after it was bought by Microsoft. Microsoft cancelled the standalone product, integrated BlueStripe into SCOM and ruined it. If AppDynamics has to choose between spending time making sure that it is a market leader in monitoring containers (Docker) or spending time integrating with the UCS, it had better choose the former.
The OpsDataStore Strategy
At OpsDataStore we believe the following things:
- The current diversity across the software and hardware stack makes it impossible for any single vendor to cover the waterfront. IBM, BMC, HP and CA failed at this 20 years ago and it is a much harder problem now.
- The pace of innovation guarantees that this will continue to be an increasingly difficult problem to solve
- Digitization means that this will continue to be a more important problem to solve
- Monitoring at each of the layers of the stack is a whole company problem. You need to be an AppDynamics, New Relic, or Dynatrace to do a good job at the application and transaction layer. You need to be an ExtraHop to do a good job with streaming analysis of wire data. You need to be a SevOne to manage networks at scale. You need to be a Virtual Instruments to really understand storage performance.
- All previous attempts to integrate across monitoring vendors have failed. Manager of Managers (MOM) failed. Alert integration (NetCool) failed.
Given the above, we believe that the only viable approach is to recognize that monitoring is going to continue to require an evolving set of best of breed vendors that specialize in particular layers of the stack.
The OpsDataStore strategy then is to leverage the investments and the expertise of the best of breed vendors in the monitoring space, and to combine their data (their metrics, objects and relationships) into a common big data back end. This is depicted in the image below.
OpsDataStore Ecosystem Strategy
We are today, proud to be partners with AppDynamics, Dynatrace, ExtraHop, VMware and Intel. We will be adding data integrations with the ecosystem above and others over time. Our strategy is to respect the positions and value of our partner vendors and to make their products more valuable to our joint customers.