Episode Summary

iland Cloud Technologist Brian Knudtson is joined by guests Ron Singler, Melissa Palmer, and Aaron Delp for a conversation about how reliability is built into a cloud platform. From the datacenters on up what does it mean to be reliable? As with life, it’s about more than just availability — and it changes over time.

Panel

Cloud Conversations

Topic 1

[03:43] Datacenter facilities are the rock upon which a reliable infrastructure is built. Should customers concern themselves with the facilities that a cloud provider uses, and if so, what should they be looking for?

Topic 2

[09:40] One often touted benefit of the cloud, particularly for SMBs, is access to a more reliable infrastructure — higher class equipment combined with better redundancy. How much of an advantage is this for customers? And is it a core consideration to the move to the cloud?

Topic 3

[18:20] How do hyperscalers, VMware-based providers, and on-premises cloud solutions differ in how they build reliability into their infrastructure?

Cloud Bites

[01:07] “When I think about reliability in the cloud, I really have to go back to the root of the word and just ask myself — can I rely on my provider to give me three things: scalability, flexibility in consumption options, and cost efficiency?” – Ron Singler

[02:09] “We need to understand the basics of infrastructure to understand how our cloud provider is giving it to us. We also need to understand how our applications are architected and deployed.” – Melissa Palmer 

[03:27] “There is a lot of old thinking out in the industry when it comes to matching the applications to the infrastructure.” – Aaron Delp 

[05:51] “It is so easy to whip out your credit card and go to a cloud provider that you just do it, right? You forget that there are all these physical components. You forget that this is someone else’s datacenter. It’s not a magic thing in the sky.” – Melissa Palmer 

[10:59] “Some people will just look at it and say, ‘the cloud is too expensive, we can do it better ourselves.’ But can you really do it better yourselves? Can you really do it better than someone who’s core business is to do it right?” – Melissa Palmer 

[11:22] “Why not just go buy a better infrastructure in the cloud instead of doing not as good yourself .” – Melissa Palmer 

[13:25] “There is a cost associated with reliability, no matter what.” – Aaron Delp 

[17:14] “Everyone’s okay with downtime until they go down.” – Ron Singler 

[20:49] “You’ve effectively put crap on crap. When you put crap on crap, well, that’s going to be the end result.” – Aaron Delp 

[22:53] “Customers wanted a cloud-type consumption model, but they also wanted to run their legacy applications in the cloud. They couldn’t re-factor all of those applications into a cloud-native format. I think that’s where the iland’s of the world show a lot of their value for customers who may have 100 or 200 or 1000 applications that need to be re-factored and they can’t afford to do that.” – Ron Singler 

[23:52] “A lot of people don’t consider a lot of these applications, the application lifecycle, will exist longer than at least one, if not two, generations of the infrastructure.” – Aaron Delp 

[26:46] “The good old spare parts depot. That is risk mitigation 101. We have a box of old parts in the datacenter in case something goes wrong.” – Melissa Palmer