Modius Critical Infrastructure Management on NonStop X

There really hasn’t been any way to miss the news of NonStop being ported to the Intel x86 architecture and with the announcement of Virtualized NonStop, the gloves have really come off. HPE is positioning NonStop as a software play and for Modius, this may just be the transformation of NonStop we have been looking to see take place for almost a decade. Data Center Infrastructure Management (DCIM) or the increasing adoption of IIoT may not come quickly to mind as mission-critical applications but with companies looking to the cloud for future capacity on demand and with initial forays into cloud computing involving hybrid configurations with considerable compute-power still present in the data center, it is even more critical than ever before. Managing every part of the mission critical infrastructure of an enterprise to ensure continuous availability of applications critical to the operational performance of the company is a requirement that demands 24×7 uptime.

The relationship between Modius and NonStop dates back a long way with a number of people coming to Modius with NonStop skills. Mark Merala, our technical lead, once worked in the Distributed Systems Management group at Tandem Computers then managed by Margo Holen and where his peers included Brad Poole of comForte and Richard Buckle of Insession and GoldenGate and now a principal at Pyalla Technologies. These associations have meant that Modius has been mindful of the value proposition that comes with NonStop and yet the history of collaboration between Modius and HPE hasn’t produced quite the results both sides had hoped for – and it really did all come back to the cost of a NonStop system. However, this may change once again as we dig deeper into the capabilities of NonStop X and specifically, Virtualized NonStop.

The first installations of DCIM were on Windows, and Windows continues to be the main platform deployed at customer sites, but back in 2009 the application was ported to NonStop and was installed at the Advanced Technology Center (ATC) that was then located in Cupertino. In follow-up commentaries and posts, it was reported that the team at Modius had been deploying their product on low cost commodity servers but, over time, they had found it necessary to utilize clusters. The complexity this added to the application was considerable, so much so that a NonStop’s architecture looked extremely beneficial to Modius. “No more issues with configuring a network of servers to ensure high availability – the inherent N+1 redundant architecture at the core of NonStop makes a very persuasive argument – an absolutely mandatory requirement where the ‘availability’ of mission critical facility is at stake,” we told HPE at the time.

But what is Modius OpenData and what does it really do? And could it still benefit from running on a NonStop system? According to one user in a quote we posted on the Modius web site, “Modius combines fully-redundant 24×7 monitoring with continuous measurement of critical equipment. Together, these capabilities allow us to drive forward our aggressive efficiency optimizations with the confidence that the integrity of our facility won’t be compromised.” It has been a while since we first ported OpenData to NonStop but with x86 support, a new opportunity may arise – clearly, if Virtualized NonStop can run on commodity x86 servers as has been reported, then this may prove to be the very catalyst we need and the industry does seem to be coming back to NonStop given the more widely distributed and ubiquitous IIoT implementations that demand scalable and resilient computational platforms

This month, Modius announced that its upgraded Power Capacity Management module for OpenData v3.7 will change the way data centers provision for power capacity. Before IoT, data centers provisioned power capacity based on the power requirements published by the manufacturer of each piece of IT equipment. But this method often led to gross overprovisioning of power. The Modius PCM Module for OpenData v3.7 helps Facility and IT personnel work together to insure that power related outages are avoided and available power is used efficiently to support high levels of equipment density in the data center. The PCM module allows Facility personnel to build an exact model of the data center power infrastructure, from the utility to the servers in the racks, augmenting the manufacturers’ specifications with real-time instrumentation for devices in the power-chain. This allows the end-user to compare power draw specification to real-world data for fine tuning capacity reserves, and simulating fail-over and redundancy scenarios to verify that equipment failures won’t cause service interruptions. (For more about this announcement check out March 3, 2017, news release, Data Center Infrastructure Management Leader Modius® announces new IoT approach for power capacity management

Should the market warm to Virtualized NonStop and enterprise users embrace Virtualized NonStop running on commodity, off-the-shelf servers (COTS), as HPE predicts then this may accelerate our own plans to once again look at the costs to port OpenData to NonStop X. We have been told that the NonStop X and Virtualized NonStop share the same APIs and indeed, the OS interacts the same way whether running on physical or virtual machines and this is something we want to take a look at. As we know more about the general availability and pricing of Virtualized NonStop then more serious consideration of porting Modius OpenData to NonStop will be given – and if this is something that appeals to the NonStop community then please let us know. We are openly soliciting feedback from the community and a positive reaction will go a long way in accelerating our timelines. You can reach me by email at craig.compiano@modius.com or simply call me at the office, 415-655-6700.

Craig M. Compiano, President Modius, Inc.
craig.compiano@modius.com | (m) 510-289-5169 | (o) 415-655-6700