Remember that a cloud operations SaaS provider should perform better than your own IT operations, because IT is their specialty. This leaves you free to focus on what you do best — make products. How can you be sure that a cloud data center and SaaS provider is providing equal or perhaps better service than your internal IT department? The provider should offer metrics that are meaningful to your CIO and other senior technical leaders.
At Plex Systems, we track metrics across four areas – availability, reliability, scalability, and security.
Some applications are more critical than others. For example, while an outage in an HR or email application is certainly an inconvenience, downtime in plant operations means products aren’t being made — and this has a direct impact on the bottom line. In manufacturing, any outage is a business outage.
Look carefully at a potential provider’s availability metrics year over year. Is the provider willing to commit to a SLA (service level agreement) of over 4 nines (99.99%) availability for their MES? This will equate to roughly 53 minutes of unplanned downtime per year. At Plex, our availability has been between 99.995 and 99.997 (13-25 minutes per year) over the past two years — well above the industry average.
How is the provider equipped to handle disasters such as a weather event or terrorist attack that might take out an entire data center? How many data centers does the provider have and how are they configured? For comparison, I can tell you that Plex recently expanded its data center footprint, moving from a single data center with a DR site to three data centers with the application extending across those sites. By migrating to an active/passive data center set-up, we have a passive site that provides more than just DR. Each site is identical so there is no degradation in service in the case of a disaster.
We have another level of protection because we use two data center providers—which means no lock-in. We use the same modular Dell EMC Vscale converged infrastructure for both and our application is optimized for that architecture. It’s easy to spin-up, so in the event of an issue with one of our providers, we can easily switch over to the other.
In addition, while many providers test failover once a year, Plex tests multiple times per month so we can stay well inside our SLAs. With the active/passive architecture, we have also been able to shrink our maintenance window for infrastructure updates. We can do updates before failover, increasing uptime. More uptime means more revenue for our customers.
Reliability can be summarized as how well the application performs its intended function. In other words, is the application working as it is supposed to? In a mission-critical manufacturing application, it is important to track all transactions – this could mean billions of transactions per day. Is the cloud provider’s infrastructure equipped to handle large transaction volumes?
In the event of a failed transaction, the provider should be able to determine why that transaction failed. For example, did it time out or was there an execution error? Can the provider link application performance to user experience? And does the provider have analytics that can determine the underlying cause of performance issues and provide proactive measures to prevent those problems?
Plex tracks, on average, 5 billion transactions per day. With that volume of transaction data, as well as the data that Plex collects from sensors, machines, people and other applications, Plex analytics will expose a problem very quickly — enabling the issue to be addressed before it affects operations.
One of the benefits of a cloud-based SaaS application is seamless on-demand scalability as your business needs grow. As you look at providers, be sure that the solution is multi-tenant so that no single customer affects any other.
We’ve already covered the availability benefits of a provider with multiple data centers. Scalability goes hand-in-hand with availability. With Plex’s three identical active data centers and modular infrastructure, customers can easily scale up or down and have plenty of room for growth — without sacrificing uptime or performance.
With respect to security, make sure the provider is keeping up with industry security and privacy standards and regulations. In a multi-tenant solution, is your data separate and secure from other customer’s data? Does the solution enable role-based segregation of data within the company? What are the encryption policies? What are the protections against cybercrime and hackers?
Plex closely follows developing standards and regulations that make sense for the business. We are already a year into the Privacy Shield data protection policy — one of the early adopters after Safe Harbor went away. We are ITAR compliant, a US government regulation that controls how and where data is housed, managed and accessed. And we met NIST 800-171 compliance at the end of 2017 — important for SaaS multi-tenant providers that provide goods to the US government.
Plex is multi-tenant and we have role-based data access that allows employees to only access data that is relevant to their department and position, while protecting the privacy of sensitive information. We offer encryption-at-rest and encryption-in-transit. Our employees receive training that ensures that they don’t accidentally export data to the wrong person or customer. And finally, we closely track security updates and patches that protect against cybercriminals.
The bottom line
So can you trust your manufacturing business to the cloud? Well that depends on the provider. Armed with the proper questions, you can easily determine whether your SaaS provider is going to meet and exceed your requirements for availability, reliability, scalability and security.