Last week Amazon outage was a splashing event for everything related to cloud computing. Amazon S3 in east cost went down and knocked out many cloud services around the world. In CAD / PLM ecosystem it was almost a silent event. Autodesk cloud infrastructure partially went down. Also GrabCAD informed customers about problem. All other PLM vendors kept radio silence.
I’ve been following multiple articles after the event providing variety of opinion about how to trust (or not) cloud infrastructure. I also been asked by some of my readers online and offline if cloud providers can be trusted and how reliable cloud platforms for the future of mission critical systems such as PLM.
My favorite article was Infoworld’s AWS outage proves one cloud isn’t enough. Thanks Stan Przybylinski for sharing it online. The main point of the article is dead simple. The idea of cloud computing is not about how to stuck your PLM server in the cloud and claim reduced IT cost and effort, but to provide a reliable fault tolerant cloud infrastructure. I like the following passage:
All of this means you need to develop a system that tolerates latency. Your fake cloudification, where you stuck SQL Server in the cloud and plonked a load balancer in front, isn’t likely to cut it without real thought and re-engineering.
And the following passage gives you a recipe in a nutshell – horizontal scale and cheap redundancy will rock the future.
In the past, companies set up multiple data centers, WAN replication, redundant DNS, clustered services … all these have been doable, albeit less affordable in the past. But today’s database technology, micro-services architecture, and vastly improved software make redundancy much cheaper and easier. But don’t get hung up on using only one provider’s offerings.
It made me think about existing PLM platforms architectures and their ability to follow new new type of architectures. You probably remember my article Will PLM microservices eat PLM dinosaurs? Existing PLM platforms are going back 15-20 years in the lifecycle of development and multiple acquisitions. These platforms are mostly monolithic application and servers relying on database architecture to store and manage data. One of the biggest problem of existing architectures is move from one product to another or to update existing products. The monolithic application nature makes that problem very painful. New cloud architectures can provide an alternative by bringing a set of microservices and replace existing product with new type of cloud agile platforms.
What is my conclusion? New PLM architectures should kill single point of failure. For many existing PLM platforms SQL driven backed is the one to kill. But it is not simple when PLM vendors are hiding Oracle or SQL server driven tenant oriented applications behind IaaS infrastructure. Highly available, redundant cloud architectures is the future of new sustainable PLM platforms. The one that can sustain next AWS outage. This is a note to PLM architects of the future. Just my thoughts…
Want to learn more about PLM? Check out my new PLM Book website.
Disclaimer: I’m co-founder and CEO of openBoM developing cloud based bill of materials and inventory management tool for manufacturing companies, hardware startups and supply chain. My opinion can be unintentionally biased.