Whereas last month we raised the topic of security, this month, it’s about what many vendors are discussing as well – leveraging our knowledge of NonStop in the work we do for other platforms. While many vendors can provide their users with server clusters, when it comes to true availability these users are left to develop scripts, insert APIs into their applications, and rely on external services like data replication to achieve the levels of availability anywhere close to what comes out of the box for NonStop users. This can be a disruptive approach, where the invasiveness can itself lead to further problems, with few users prepared to do the routine take downs to ensure everything goes to plan.
In an upcoming column for The Connection, Pyalla Technology’s CEO, Richard Buckle, wrote of how he wasn’t suggesting “that the LUW package itself can be made fault tolerant, but rather, there’s structural changes to the underpinning infrastructure and framework that can be made by those with a working knowledge of NonStop that improves the availability of their applications on LUW systems.” Furthermore, said Buckle, for NonStop vendors to take these steps, it “is a wise business decision for them to take (as) I would rather have a healthy population of vendors deriving income from multiple system offerings than a weak population of NonStop vendors barely getting by.” Look for his complete story in the column Back for more … that will appear in the Mar – Apr, 2015 edition of The Connection.
DataExpress NonStop (DXNS) implicitly enjoys the system/power/network etc. redundancy that comes with residing on the NonStop platform. DataExpress Open Platform (DXOP) primarily lived on a windows platform; should Windows become strangled – down went your primary FT (file transfer) platform. That was unacceptable to DataExpress and because we had a big brother version that would never fail, it constantly prodded us to think of how to provide the same resilience to the Open platform. What we have come up with is a far more robust offering that would have been difficult to develop had we not enjoyed the NonStop experience! Are we 100% redundant? I think there will be NonStop traditionalists who will always pick holes in what we have accomplished, but at DataExpresss, we believe it would be. And quite obviously, whereas Windows was used as an example above, there is now a common code base across Linux, Unix and Windows (LUW) and we are even looking at OSS (for NonStop X) in the near future.
The NonStop vendor community isn’t alone with leveraging NonStop capabilities on platforms apart from NonStop as we have heard at recent conferences from parties close to NonStop that HP too is doing something similar. With the commoditization of NonStop nearing completion in the NonStop X systems, sharing the underpinning technology with other development groups within the HP Mission Critical teams seems to suggest that NonStop will have an impact on all of them. As hybrid computers continue to be referenced as being viable solutions to business problems and with NonStop being a part of the proposed HP hybrids, for all vendors, as Buckle noted in his column, “experience on systems apart from NonStop will become very important.”