July 10, 2018 at 5:58 am
Comments posted to this topic are about the item Serverless Code
July 10, 2018 at 7:50 am
Will be interesting to see where this goes in the next 10 years. IMHO Azure SQL and other cloud offerings are a better fit for smaller companies. Once you get too big, it is probably advantageous to roll your own. I am not surprised to see the pricing strategy to be complicated and obfuscated. My wager is that in years time the key benefit will not be lower cost of cloud resources but rather speed and no startup costs for infrastructure and staff.
July 10, 2018 at 8:31 am
I think it's there now (speed and flexibility over cost), but what we're catching up with is trying to compare some performance on-premises to a cloud. In many cases people want a predicable performance because they get that often locally. They don't, and many people don't want to accept variation, so they overbuy.
I do think costs are a little high. Now they're roughly on a 2 year payback, and I think they need to move closer to a 3 year frame.
The other thing that the cloud needs is a moving set of workloads, not a constant one. The more constant, the more you can think about doing on premise, though it's not simple. Today lots of companies have an investment in a data center, so moving to the cloud makes less sense. If I've built (or rented) xx sq ft of space, and am paying people, power, cooling, what does the cloud get me? Not a lot. In the future, if I can get rid of those payments, maybe the cloud makes more sense. I think the IaaS stuff is really catching on as contracts expire and companies consider leaving current spaces, or can sell them
July 10, 2018 at 2:15 pm
I work for a company that has 20 offices scattered throughout the US/UK. Application performance is HUGE, Security is HUGE and downtime is not an option. Our network team laughs about "The Cloud" and the costs. Also the integration between the applications is also HUGE. Don't ever see us moving to the cloud...
July 10, 2018 at 6:37 pm
Jeff Mlakar - Tuesday, July 10, 2018 7:50 AMWill be interesting to see where this goes in the next 10 years. IMHO Azure SQL and other cloud offerings are a better fit for smaller companies. Once you get too big, it is probably advantageous to roll your own. I am not surprised to see the pricing strategy to be complicated and obfuscated. My wager is that in years time the key benefit will not be lower cost of cloud resources but rather speed and no startup costs for infrastructure and staff.
Nope, not at all. Once you get into those enterprise agreements, you can save a great deal. Ends up becoming very self-service without any limitations on hardware or licensing costs. You get a budget, you spend what you want within that budget and it's all controlled within the same tools you are using.
There certainly is dangers here. Plenty of large organizations have been caught like a kid in a candy store. In meaning, it's very easy to overspend, over integrate, and have very knee-jerk reactions that lead to security risks. But like anything, having a good business process and plan goes a long way.
July 13, 2018 at 4:29 pm
Chris Hurlbut - Tuesday, July 10, 2018 2:15 PMI work for a company that has 20 offices scattered throughout the US/UK. Application performance is HUGE, Security is HUGE and downtime is not an option. Our network team laughs about "The Cloud" and the costs. Also the integration between the applications is also HUGE. Don't ever see us moving to the cloud...
I would agree there are many circumstances were the cloud does not work. But I think the cloud maybe best as a Disaster Recovery Site, not necessary a full mirror of all servers, but enough to keep the business running when it hits the fan.
But like Steve said, it is a fantastic option for new companies without the know it all and finances to build their own systems.
July 16, 2018 at 4:34 am
I think the key to getting the best from serverless computing is in the design and architecture of the application. The costs act as a spur to getting it right.
Looking at transaction history do you see ebbs and flows in traffic either seasonally or hourly? If you have a well designed asynchronous, distributed application then serverless computing can save you a huge amount of money.
I see the transparency and visibility of the costs as a major advantage. Having to over-provision to cope with peak load estimates for 3-5 years in the future is expensive and also doomed to fail. A DB server geared up to cope with 3-5 years growth in traffic will have other apps and traffic shoe-horned onto it because it is an expensive resource with plenty of resource in year 1 but will run out of steam before its projected lifespan ends.
One of the problems with a serverless architecture is working out how to test it in an automated fashion. How do you unit test a serverless function?
July 16, 2018 at 7:35 am
David.Poole - Monday, July 16, 2018 4:34 AMOne of the problems with a serverless architecture is working out how to test it in an automated fashion. How do you unit test a serverless function?
Since you can deploy these with code, I'd assume you'd have a matching "test" serverless function that you treat as a black box and observe ins/outs. Of course, this means you're counting on things like connections configuration that changes between test and prod being done correctly.
Maybe there are emulators as well. If not, testing is certainly going to be hard.
Viewing 8 posts - 1 through 7 (of 7 total)
You must be logged in to reply to this topic. Login to reply