TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

MVPs and $100k AWS Bills: Reflections on our launch

94 点作者 paulstovell超过 5 年前

15 条评论

pjc50超过 5 年前
&gt; Octopus Cloud customers could start a free 30-day trial, which meant that those hundreds of trial signups per month, each of which cost us $100 to host, quickly added up.<p>&gt; We also didn’t have our pricing quite right. We initially launched Octopus Cloud at $10&#x2F;month, with a different pricing model from the one we currently use. Unfortunately, this was one of the most painful lessons we learned because the deficit between what we were charging and spending was magnified by the sheer number of people using Octopus Cloud; continued growth would further amplify the problem.<p>&quot;Every person who fills in a web form gets to burn a $100 bill of ours&quot; is the real disaster here.<p>I know it&#x27;s trendy to think that software isn&#x27;t like regular business, and you can lose money but make it up on volume, but going to a cloud vendor gives you a real cost of goods sold, and that needs to be accounted for.<p>(also, the irony of a company whose product appears to be about cloud deployment having a disaster by deploying to the cloud should not be lost on us)
评论 #21207134 未加载
评论 #21202909 未加载
评论 #21207301 未加载
评论 #21202975 未加载
teddyuk超过 5 年前
&quot;To bring Octopus Cloud to market quickly, we did the simplest thing possible; we took our self-hosted Octopus Server product and bundled it into an EC2 instance for each customer that signed up. We had to make changes to the product, but mostly around permissions.&quot;<p>I have used octopus and I could have told them that they can&#x27;t lift-and-shift. The amount of requests the app makes is ridiculous, the way they wrote their database is appalling (actually crazily dumb <a href="https:&#x2F;&#x2F;octopus.com&#x2F;blog&#x2F;sql-as-document-store" rel="nofollow">https:&#x2F;&#x2F;octopus.com&#x2F;blog&#x2F;sql-as-document-store</a>). Everything about it would mean they can&#x27;t scale in the cloud without wasting a ton of money.<p>&quot;Cloud stuff can be really expensive&quot; - yes if you write dog shit, putting that in the cloud will be expensive.
评论 #21202657 未加载
评论 #21209603 未加载
评论 #21202894 未加载
评论 #21207598 未加载
echelon超过 5 年前
I have a streaming multimedia app that I want to launch. It ingests from the client, does some transformations, then sends it back to them (or to other parties) in realtime. It&#x27;s pretty novel though - Snapchat adjacent.<p>I can&#x27;t do it clientside. Nature of the problem.<p>I&#x27;m writing it in Rust, so it&#x27;s relatively fast. But throwing thousands of people at it will still require extreme scale. I&#x27;m going to use K8S horizontal autoscaling. This will be <i>obscenely expensive</i>.<p>I&#x27;m not sure I can afford this. But I think a month of successful demonstration will get me funded (or bought out). So <i>in theory</i> my costs could wind up getting paid. Maybe.<p>Do I risk launching this and potentially having $100k multiples of AWS bills? I can&#x27;t afford that. I don&#x27;t want to deplete my life savings or potentially go bankrupt.<p>Could I put it on a corporate card as an LLC and do my best to make it work? Then fold without harm or fowl if it doesn&#x27;t? Is that unethical? Could I get sued?<p>All that said, I&#x27;m almost certain that a successful demo at scale with the Internet thrown at it will get me money. It&#x27;s a really cool and novel product. People will be talking about it.<p>What should I do? Advice will be very appreciated.
评论 #21202935 未加载
评论 #21202892 未加载
评论 #21202924 未加载
评论 #21217632 未加载
评论 #21202899 未加载
评论 #21209627 未加载
评论 #21202861 未加载
评论 #21202808 未加载
blantonl超过 5 年前
The architectural decision to deploy each customer on their own segmented components to make sure that &quot;no one user’s data to mingle with another&quot; seems to me like a solution in search of a problem.<p>First, the deployment was on AWS (EC2 and RDS among other services) which was probably already co-mingled with other AWS customers.<p>Second, when you are selling a SAAS offering, the assumption is <i>already there</i> that you&#x27;ll build security and multi-tenancy into the app and the underlying infrastructure shouldn&#x27;t matter.<p>And 100K&#x2F;month for a MVP SAAS offering. I&#x27;m speechless. I just, I can&#x27;t even understand how this burn rate passed muster with the leadership.<p>But Kudos to the team for owning up to it and sharing. Lots of lessons learned here.
评论 #21207793 未加载
评论 #21203070 未加载
scarface74超过 5 年前
The lesson to be learned here, and the one they admittedly already knew in advance, is if you do a lift and shift to AWS without changing your processes to be cloud native, you will always spend more.<p>Unfortunately, I’ve met way too many old school netops guys who got one AWS certification and all they knew was how to log into the web console and duplicate their colo infrastructure, charge their client a lot of money, and call it a day.<p>That being said, I’ve used a lot of CI&#x2F;CD tools and OctopusDeploy is by far my favorite.<p>Even though my company is very much a “pay someone to do the undifferentiated heavy lifting” type of company, OctopusDeploy is so easy to use and maintain, I am not sure we would migrate.
评论 #21202773 未加载
remon超过 5 年前
I&#x27;m sorry but this was a bit of a painful read. Hindsight is 20&#x2F;20 but this was pretty poor engineering and business modelling from the start.<p>&quot;To ensure there was no way for one user’s data to mingle with another, each cloud instance had its own dedicated VM, database, and a large number of security configurations to prevent any funny business.&quot;. AWS has a myriad of security and isolation tools to do that in significantly better and cost efficient ways.<p>&quot;Octopus Cloud customers could start a free 30-day trial, which meant that those hundreds of trial signups per month, each of which cost us $100 to host, quickly added up.&quot; So you&#x27;re essentially allowing anyone to press a &quot;Submit&quot; button to spend $100 for you. This sort of below-cost methods to gain marketshare works for Amazon, Google, Microsoft and son because they have the deep pockets but it&#x27;s a questionable strategy for smaller companies.
rubbingalcohol超过 5 年前
This is insane. It&#x27;s the engineering equivalent of smashing one of your hands with a hammer and then being surprised it hurts. Well no shit.<p>At that level, at the very least, it would make sense to run some amount of physical infrastructure.
notyourday超过 5 年前
Pardon me for asking a really dumb question: but do you have a positive item charge - CGS? Because I&#x27;m sure one can get a great demand by selling a $1.00 for $0.90 but that&#x27;s not a business model.
评论 #21207753 未加载
评论 #21207339 未加载
mothsonasloth超过 5 年前
Can&#x27;t wait for this Cloud meme to die out and we can go back to On Prem but instead of VMs managed by VMWare, Cloudstack, Vagrant, LXC etc, we will have nodes managed by Kubernetes.
评论 #21203026 未加载
评论 #21203233 未加载
DVassallo超过 5 年前
&gt; We quickly learned that everything in AWS has some kind of service limit<p>Except S3. Unlimited objects, and unlimited Put and Get requests.
评论 #21206487 未加载
评论 #21206127 未加载
poxrud超过 5 年前
<i>To ensure there was no way for one user’s data to mingle with another, each cloud instance had its own dedicated VM, database</i><p>No this will not protect you. Understanding the proper way to configure IAM roles with principle of least privilege will.<p>The author mentions that <i>The issue here is Octopus is like a CI server - customers can run arbitrary code in the instance (not a problem most SaaS apps have). So a VM per customer was almost the bare minimum.</i><p>If you&#x27;re using the same IAM role for your EC2 instances, and you&#x27;re allowing your customers to run arbitrary code, well now at the very least they have access to everyone&#x27;s databases.
TedLePoireau超过 5 年前
Rent dedicated server from OVH, install ESXi, hire a sysadmin. You will save huge amount of money. Cloud can be great, but sometime it looks like people forget there are other solutions.
评论 #21210429 未加载
Havoc超过 5 年前
Well glad they had the capacity to recover.<p>Even accounting for hindsight being 20&#x2F;20...every customer gets their own VM is a rather ambitious approach to scaling.<p>Would love to know what made them shift to Azure. Especially while embracing linux at same time.
评论 #21208760 未加载
plasma超过 5 年前
Octopus is a fantastic deployment product for .NET (happy user here).
评论 #21213861 未加载
评论 #21212686 未加载
评论 #21202992 未加载
omouse超过 5 年前
Ha, I&#x27;m re-reading The Lean Startup and this paragraph seems like a contradiction in terms of what an MVP is:<p><i>We decided to build an MVP based on our best estimates and test the market that way. The goal was to launch something in 6 months and test if the demand was there; and if it wasn’t, we’d only wasted 6 months. We chose to optimize for getting to market quickly rather than worrying about how much it would cost.</i><p>Contrast it to what the Lean Startup principles say:<p><i>The Lean Startup methodology has as a premise that every startup is a grand experiment that attempts to answer a question. The question is not &quot;Can this product be built?&quot; Instead, the questions are &quot;Should this product be built?&quot; and &quot;Can we build a sustainable business around this set of products and services?&quot;</i><p>So instead of doing what a lean startup would do, they did the <i>opposite</i>. Let&#x27;s see what they could have done instead...<p>&quot;How many customers will actually use it?&quot;<p>- Lean startup: create a signup page with a mock up or landing page and a few buttons that work and every potential behind a &quot;coming soon&quot; screen - Their approach: actually build the whole product and then launch it and see who&#x27;s interested<p>&quot;What is it all going to cost?&quot; - Lean startup: As little as possible to discover the insights we need to iterate and evolve the product - Their approach: well we already paid the salary for our engineers and we&#x27;ve given them six months, so what&#x27;s half the salary+benefits of X number of engineers?<p>&quot;What should we charge for it? Is it going to cover the infrastructure costs?&quot; - Lean startup: experiment on costing and pricing as you go along - Their approach: give away the product for a 30-day trial at a high cost to ourselves! Unbelievably expensive customer acquisition costs.<p>Reference: <a href="http:&#x2F;&#x2F;theleanstartup.com&#x2F;principles" rel="nofollow">http:&#x2F;&#x2F;theleanstartup.com&#x2F;principles</a><p>And now I guess I&#x27;ve truly learned why product managers exist. Someone let this crazy experiment run for <i>six months</i>. You&#x27;re saying in six months you couldn&#x27;t spend one month finding the simplest&#x2F;cheapest ways to test these various hypothesis?<p>$100k customer acquisition spend is insane.<p>The whole post is just a lesson in contradicting what an MVP <i>minimal viable</i> product is.<p><i>Even when we were building v1, the team knew it wasn’t the ideal architecture. Before v1 even launched, there were plenty of conversations in our Slack about whether we should port Octopus to Linux and run it on Kubernetes, or see if we could run it on Windows within Kubernetes or use Nomad by Hashicorp?</i><p>Lol, engineers always want something fun to work on and sometimes it turns out to be a great idea. But they porting software for an MVP? Why not evolve the new product so that it generates more revenue or reduce the costs in some other way instead of talking about porting and rewriting?<p>If you&#x27;re at an established company that has money to burn, go ahead, read the whole blog series. If you&#x27;re trying to create a new startup, whether it&#x27;s a product or service, avoid this article or read it as a warning. When you&#x27;ve spent as much as a decent engineer costs in <i>one month</i>...something&#x27;s gone off the rails.