Organisations today are facing numerous pressures to enable a remote workforce, particularly in the IT function, since we have entered the post-Covid era. At a time when the traditional modus operandi is constantly being tested, there are some ‘new’ approaches that have actually been in use in other parts of the market for a while now.
We can take several lessons from the consumer tech world and how it leverages automation and AI to reduce maintenance and ease automation. Let’s take at the Nest thermostat as an example. A single thermostat changes temperature about 1500 times per year, so a large house with 3 thermostats changes temperature about 5000 times per year. In fact, it is the Nest cloud-based control plane that enables numerous capabilities beyond keeping your house at just the right temperature:
- Thermostat software is always up to date;
- It provides remote monitoring and alerts and supports remote administration to respond as needed;
- It allows simple automation of your HVAC with home automation tools.
So while the end point is really important — that is, it has to tick the boxes to be considered as a replacement for a traditional thermostat — the real magic is in the features enabled in the cloud-based control plane. If you could apply this cloud-based control plane model to your data centre and edge storage infrastructure, what would that look like?
Tip #1: Seek out storage infrastructure equipped with Gmail or Tesla-like storage software updates.
As we all know, updating software on your external storage infrastructure is a huge headache, and can take weeks or months to propagate an update across a fleet of devices. And just when you finish, you get a new update from your vendor and it starts all over again. It’s a little like painting the golden gate bridge — once you get to the end, it’s time to start all over again.
How to avoid cloud vendor lock-in and take advantage of multi-vendor sourcing options
It turns out that as much as 75% of the code delivered in storage array software updates are related to management, with just 25% related to the IO path. What this means is by leveraging a cloud-based control plane for your on prem storage, 75% of the updates just happen in the cloud as they are released. New features simply show up on your UI, much like your Gmail, or for those of you lucky enough to cruise around in a Tesla.
The IO path has a much smaller footprint, and can be distributed in small, less frequent and non-disruptive micro-updates. With these, you might first do a small update in your lab and assuming everything checks out, you push it to hundreds of endpoints with a single click.
Tip #2: Pivot from traditional storage management and monitoring to SaaS-based storage operations-as-a-service.
A number of storage platforms in the market today provide some level of cloud-based monitoring, but administration is typically done in a completely separate on-prem console. Alerts and troubleshooting are great, but co-locating AIOps and storage administration means you can have analytics & alerts in the same cloud that provides administration. That means immediate, even automated, action is possible on predicted current and future issues. Beyond maintenance, a unified AIOps administration approach lends itself to providing proactive recommendations to optimise performance, keep users happy, improve utilisation and ultimately spend less on capacity.
Delivering collective intelligence with the five pillars of AIOps
Tip #3: Demand simple programmability with a single endpoint in the cloud to automate your entire infrastructure
Make sure you have a single API-endpoint in the cloud to enumerate & automate all of your storage assets on-prem. Having a cloud-managed platform provides the visibility and orchestration of your assets across sites, servers and applications and you can take advantage of a single API in the cloud to then automate all or a portion of those as needed. You get an aggregated view, or you can filter by data centre or application, server group, etc. Then ask interesting questions like, where is there available capacity for a new project? What does my VMware utilisation look like? Which of my container deployments is performing best?
Generation Data: the future of cloud era leaders
Perhaps the most important attribute of this approach is the fleet management benefits. If you have 1000 servers each with 10 SSDs and a critical SSD patch in needed, how would you handle that? A cloud-based control plane will allow you to roll out firmware updates across 10,000 SSDs with a single click.
By leveraging automation and AI, IT organisations can help ease the burden of their remote workforce with automated software updates, storage operations-as-a-service and simpler programmability for at-scale automation. With the right tools, the remote IT workforce might just be something that lives well beyond our current challenging times.
