We’ve run the full range of services in the cloud here at SSWUG.org, here are some of the wins we’ve seen with our own systems. I’ll take a look at some of the pros and cons over the next couple of days based solely on our own experiences.
Our experience is based on moving from on-premise to co-located serves to cloud-based services to software as a platform. We’ve had the interesting fortune of running through it all.
Some of this may seem odd given the recent discussions here, but it’s based on our own usage and epxeriences.
Performance
Our experience with the cloud has been one of better performance overall. More consistent performance has been a surprising benefit only because our co-location provider would suffer from a maxed out server cabinet at times as we stream events and other types of things. Not tomention other customers on the same “trunks” that could impair performance.
With the cloud, we’ve seen more control, ironically, by allowing us to move between availability zones and geographic locations. Overall, with some early exceptions, it’s been a very positive environment for us. Where we experience performance issues were only when we were experimenting with database as a service – due to a number of factors, we were being throttled. This response to high activity was not something you’d necessarily see if you were on your own (or VM) system. It would “just” slow down. Once addressed, these issues have been worked out. It was a combination of bigger instance sizes and better management of resources.
Security
The security tools we have at our disposal are vastly improved when compared to on-premise or co-location facilities. Sure, we could have implemented additional things in those environments, but with the cloud provider, they are available by almost literally “turning them on” or tuning them. Our ability to control access, manage ingress and egress and all that that entails, vastly improved relative to the amount of time we had to spend working on it, and the functionality provided.
What’s more, we had very good experiences applying best practices and working with the provider to make sure we were doing it right. The support was very helpful.
Hardware Recovery
Sort of a two-edged sword here. We have had hardware failures from time to time (it happens), but with our own equipment, or co-located servers, it would have resulted in surprise bills for hardware acquisition and configuration services.
With the cloud, it’s literally as simple as deploying a new instance, often just a few clicks of the mouse, and dropping it behind load balancers and such. We took great care to document our systems and deployment processes, so the recovery process was a relatively minor tedious inconvenience as we walked through our docs to deploy and recover.
Support
Support has been nothing short of outstanding. Architecture, best practices, practical application, how-to and all of those – the assistance is there. Yes, it can be frustrating (it can take 12 to as much as 72 hours in our experience) so it’s defintely not like walking down the hall and asking questions. But the support has been complete, insightful and spot-on. We’ve learned a great deal from our support interactions with the provider.
We’ve experienced signficiant positives. I’ll go into some other items in the day(s) ahead – positive and negative – that we’ve seen. No solution is perfect of course. There are good and bad elements no matter which direction (or combination of directions) you choose to go for a given solution.
Do any of our experiences ring true for what you’ve seen?