4 ms·
I agree that scalability is rarely an appropriate reason to go in on AWS; however, “managed services” is a really good reason. You don’t have to operate your ow
by throwaway894345 6y ago
I agree that scalability is rarely an appropriate reason to go in on AWS; however, “managed services” is a really good reason. You don’t have to operate your own equivalent services, but rather you pay for Amazon engineers to operate them on your behalf. Of course, there are some people for whom running their own services is the more appropriate scenario, but I think this is not a good default for anything more complex than a Rails app.
Someone will snarkily mention that Amazon services go down from time to time across an availability zone or even a whole region, and presumably this never happens on prem (I guess the idea is that Amazon engineers are well-below average?) simply because you rarely hear about it when $MomAndPopCRUDShop goes down. On that note, in my experience, customers are really sympathetic when an Amazon outage brings you down (because half of the Internet is also down) but not very sympathetic when it’s just your site that’s down.
- ocdtrekkie 6y agoIt's not that Amazon has worse engineers: They have great engineers. The difference is they are Amazon's engineers, not your engineers. My IT team is not going to break something on the most important day of the month for my business. Sometimes even when I want to do something mundane on the network, it's "eh, let's wait until not (insert key date here), just in case". Amazon does not care what day is important to my business, they're Amazon and they do what they want. You are going to pay IT staff either way, if it's your own hires (or an MSP), or you are paying via AWS fees. If you're going to pay for IT, why wouldn't you pay for IT that takes orders from you and cares about your business? The other thing is, Amazon's maintenance and upgrades is based around Amazon's need to remain competitive and turn a profit, and to support businesses other than yours which may have larger needs. (The rollout of a new feature that causes an outage for you might not have any benefit to you anyways.) Similarly, a single patched Exchange server running on a single VM on a VM host with a UPS backup on it is generally speaking, more reliable and has better uptime than Office 365. Hilariously, Office 365 also costs a lot more.
- throwaway894345 6y ago> You are going to pay IT staff either way, if it's your own hires (or an MSP), or you are paying via AWS fees. If you're going to pay for IT, why wouldn't you pay for IT that takes orders from you and cares about your business? Because my IT staff budget would go up if they have all of the additional responsibilities that we currently outsource to Amazon by considerably more than what we would save on our AWS bill by doing it in house. This makes sense because Amazon has a lot more customers besides just me paying for their engineers to manage a given service. So I would be paying a premium "for IT that takes orders from me and cares about my business" and that premium is rarely worthwhile. In the worst case, I have an important customer demo and an Amazon outage occurs--I just explain to my customer that we're affected by the same Amazon outage that's affecting all of their other vendors and half of the rest of the Internet and I forward them the Amazon post-mortem. This has never been a problem for me to date, and I doubt there are many people who have lost deals on account of Amazon in excess of their savings for using Amazon (no doubt there are a few companies for whom this doesn't hold and they should seriously consider on prem). > Similarly, a single patched Exchange server running on a single VM on a VM host with a UPS backup on it is generally speaking, more reliable and has better uptime than Office 365. Hilariously, Office 365 also costs a lot more. If I had claimed that every managed service was worth its price for every user, then you'd have successfully refuted me. :) No doubt there are some managed services that aren't worth the price, and even an individual managed AWS service probably doesn't justify the cost of integrating with AWS; however, if you're going to be needing analogs for IAM, Lambda, EKS, SQS, CloudWatch, etc, you're probably going to be spending more on balance (and incurring more downtime) by doing it in house (consider also the difficulty in finding engineers with experience in your exact matrix of tech choices versus general AWS experience).
- sokoloff 6y agoO365 had a couple of bad years, earning the nickname "Office 360" for a while. This year, I can only recall a single outage and we've been overall happy with it. O365 costs more than an Exchange server only in part because it comes bundled with all of MS-Office.
- ocdtrekkie 6y agoIf you bundle Exchange + CALs + Office Professional Plus, over a three-year version upgrade cycle, versus three years of Office 365, Office 365 costs about twice as much, last I checked. And, since Microsoft supports it's software for ten years, you can skip one or two versions of Exchange/Office and still get security updates, and pay a quarter or a sixth of what it costs to go with Office 365.
- fomine3 6y agoFor comparing email service, cloud services are superior for spam reputation because they are big.
- ocdtrekkie 6y agoSure, but the problem with this is often overblown. If I was going to send mass mail, I might prefer to use an outside service like Sendgrid or Constant Contact to avoid that risk, but for general email, it isn't a big deal to operate your own on-premise server.
- sokoloff 6y agoI run my own email domain for my family. We have DKIM and SPF setup. We've been on a clean, constant IP for 5+ years. We still have deliverability problems that I don't have from O365 (work) or gmail. We probably have even more than we detect, because these are typically invisible until you can "discover the absence of a signal".
- cutler 6y agoThat Rails app might just be the next Stripe, Shopify or AirBnB.
- throwaway894345 6y agoI’m not sure what point you’re making exactly, but Stripe and AirBnB (not sure about Shopify) are all on AWS. > A year after Airbnb launched, the company decided to migrate nearly all of its cloud computing functions to Amazon Web Services (AWS) because of service administration challenges experienced with its original provider. Nathan Blecharczyk, Co-founder & CTO of Airbnb says, “Initially, the appeal of AWS was the ease of managing and customizing the stack. It was great to be able to ramp up more servers without having to contact anyone and without having minimum usage commitments. As our company continued to grow, so did our reliance on the AWS cloud and now, we’ve adopted almost all of the features AWS provides. AWS is the easy answer for any Internet business that wants to scale to the next level.” > Since 2011, Stripe has delivered its PCI-compliant payment platform entirely on AWS, relying on the security best practices as well as easy auditability of the AWS platform. Stripe wants to make it easier than ever for developers to process payments on their web and mobile applications. Using AWS provides Stripe with access to a world-class infrastructure that helps it scale seamlessly and increase developer productivity.
- cutler 6y agoI was referring more to the throw-away reference to Rails apps being small potatoes.
- throwaway894345 6y agoI wasn’t suggesting they were small potatoes, but if your whole infrastructure is just a single rails app and you don’t need S3 or SQS or Lambda or any of the other AWS offerings then you probably aren’t benefiting much from AWS.
- busterarm 6y ago> and presumably this never happens on prem Heh, let me tell you about the time a truck drove into the sidewall of one of our datacenters and took out _both_ power sources to the facility...