Digital Ocean’s journey from TechStars reject to cloud-hosting darling | Hacker News Reader

Digital Ocean’s journey from TechStars reject to cloud-hosting darling

techcrunch.com

94 pointsbjenik56 comments

Every day, working with customers' products or witnessing stories such as DO's I realise that success is 90% marketing and 10% actual technical stuff.

You don't have to have the greatest features or too many of them or even all of them working as advertised. You don't need to take security very seriously (hello DO, WhatsApp!) or use buzz-stuff such as OpenStack (hello Rackspace, HPCloud!) or any existing "cloud standards" in order to succeed.

You just need to market it and promote it right.

I'm not even sarcastic or anything. As a techie I always frickin overlook it and this is the single most important element that can help you make serious money/business.

There's a whole bunch of people out there who don't know shit or don't really care about how great your product is technically (or not ;> ). All you need is it satisfies some basic "need" and a shiny packaging, it's enough to go out there and get some milk flowing in.

Kudos to the DO folk!

Alternatively: all the ways that it could fail don't matter if there's at least one way it could succeed.

I think at the end it came down to price. No one else was offering SSDs at the time for that cheap. Sure you get what you pay for but for a lot of people I guess it's good enough.

Yes.because it cheap and i want to test ssd. I want to test how include or require in php can be speed up.

> or use buzz-stuff such as OpenStack (hello Rackspace, HPCloud!)

Maybe I misread the tone of this post or I'm fuzzy on the history, but it sounds like you're saying Rackspace is trying to be hip by jumping on the hype-train. My understanding is that Rackspace co-created [1] Openstack and let it out into the wild.

[1] http://en.wikipedia.org/wiki/OpenStack#History

What I meant is Rackspace (also RedHat, HP, Mirantis etc) would have you believe that if you're not running Openstack then your "cloud" is shit and has no future. DO clearly demonstrated otherwise.

They're inflating the hype because they generate $$$ this way - from their public cloud, private deployments, training, certifications etc etc. Nothing wrong with that I guess, it's business.

I actually think they are right. If you are not AWS (Google or Microsoft), and you are not embracing OpenStack, you are an also ran and you will be irrelevant.

Does Digital Ocean's "cloud" offer all the same things that OpenStack does? I've been looking for another file storage provider (like S3/CloudFiles/Azure Files) but wasn't aware there were any other options than OpenStack based.

Joyent has one as well (not OpenStack-based), with an interesting "integrated compute" option where you can send code to run over stored objects without spinning up an explicit compute instance: http://www.joyent.com/products/manta

I don't think DO offers any kind of a-la-carte cloud storage, though.

OpenStack is working on a "send the code to the data" thing too. That's what Rackspace bought ZeroVM for.

If you want to try very alpha code you can kinda get it working yourself now (I think - based on mailing list discussion anyway)

Check https://www.greenqloud.com/storageqloud/

They run Cloudstack for compute, but not sure what they run for storage.

I disagree with your post in general but was particularly confused by this point. Aren't openstack and buzzyness exactly the thing you're deriding?

And I disagree that DO is missing technical chops.

Agree with your assertion here. We've been approached by a couple of vendors to try and sell us OpenStack but it doesn't really buy us much.

DO is what we need, except self hosted and supporting windows.

Totally agree.

People who are going to argue against you are going to say "But what about products that are so great that they go viral, people share them and your customer acquisition goes through the roof!". This is marketing. Making frictionless experiences that satisfy needs very quickly is a great way to market your product. Note - Many of these experiences are frictionless because the security sucks. Security is accessibility's worst nightmare.

WhatsApp did this by making it easy for my parents living in the US to download the app, add my UK phone number and voila, I can IM them for free.

DO did this by making the interface on a VPS provider "not suck" and the pricing in line with all of the junk shared providers out there.

What I've noticed about "Bubble 2.0" is that making great, frictionless products is the new marketing. It is still marketing, and I feel it's best that people recognize it then pretend it simply doesn't exist.

If this kind of "marketing" is truly the thing you overlook when making things, are you making valuable tools that are hard or frustrating to use? What are the parts of your products that have nothing to do with the friction in using them?

I think you're kind of wrong here. DO's biggest thing for me is that it's incredibly easy to use while providing all the baseline features I actually need for casual hosting. I guess you could count this as 'marketing', but I think there's a strong technical element to knowing your market and how to make what they want to do as easy as possible. DNSimple occupies a very similar niche.

In fact, part of the appeal may be that they don't use openstack, because openstack implementations to date have all been quite limited and complicated so far.

Yes. Two parts of business: engineering and marketing.

Two parts of marketing: promotion and meeting needs.

The key "marketing" of DO was:

  Existing providers offered pricing plans that were too complicated and platforms
  that weren’t always meeting the needs of their users.

It's not enough to make a "better" product, technically. You have to make a product that better meets user needs (relative to existing alternatives).

For me, it's all the pricing-reputability ratio. They've always seemed like a stand-up company, while offering a comparable product to Linode and similar...at a fraction of the price.

I jumped over to them around the time NewsBlur wrote that post about scaling to meet the new demand when Google Reader shut down. I had never heard of DO before that, checked them out, and made a the switch after a bit of research.

That seems so now a days. Folks like easy things. Plain and simple.

I use DO as a private little always-on server, I don't actually host anything worthwhile with them.

I signed up mostly because of the $5/month price and the $10 credit. That's half of what I pay for my main hosting account (over at WebFaction) and year, is about what a nice meal out costs. It's nice, quick enough for all my needs and took minutes to set up.

Yeah, I guess the line between marketing and technical is not very clear all the time.

Re Openstack, indeed it has failed the "VPS" market which is still going strong.

In my case DO succeeded more through the product than the marketing. I signed up after reading several 'who's the best hosting service' discussions and the consensus was DO was the best for small developer stuff, Linode and AWS being sometimes better for larger stuff. I've never seen an ad for DO. It's all been user recommendation.

I think he way you are explaining it may give the wrong impression that the technical aspects are easy or unimportant which couldn't be further from the truth. In fact the technical aspects (and a compelling offering) are difficult and critical but merely a baseline to get in the game. Then, all being equal, the winners are those that figure out how to gain customers (marketing, for lack of a better term).

Article doesn't disclose that Crunchfund has a stake in DO.

Totally misleading linkbait title.

Sure they were "rejected" by NYC (for valid reasons - apparently treated very well) but then went into the Boulder program after applying. And no doubt the reference given by Tisch must have certainly helped I'm guessing.

I know that TechCrunch isn't interested in discussing any actually useful details of what made DO successful...but for me it came down to them doubling the RAM per instance and using SSDs in 01/2013, creating a collection of excellent sysadmin articles, and cheap/free backups and images.

I used slicehost, though when it became rackspace cloud it became way more complicated. The main draw for me was the $5/month pricing combined with a $50 voucher they gave for black friday and they use SSDs. Even with all those benefits there are still switching costs due to having to re-setup on the new host, though I imagine most of their growth is people starting new projects.

My opinion. I am a DO user and I don't think you can use it for production application. Casual hosting is fine.

For production servers, deterministic backups are very important. DO does not provide that. I faced a lot of issues when transitioning my server from non-backup to backup mode. To do that, you need to take a snapshot(server shutdown is required for this)of server. Then, you need to destroy it to get the same ip. There is a chance that your ip may go to someone else & any services that are dependent on ip may need changes. I was lucky to loose the ip for a while & get it back after some trial & error.

Also, you don't have any control on the backup. It is taken by the system and is supposed to be anywhere between 2-3 days.

Are those the only backups you take? I have all of my code stored in git, and my database (and assorted files) get backed up each night to aws. I figure a snapshot is my first line of defense, and aws is my backup in case this doesn't work.

DigitalOcean is not a cloud or anything that resembles what a "cloud" is.

If you consider "cloud" hosting backup-enabled virtual private servers with multiple operating-system choices and multiple locations, well, hate to break it to you, but that's has been around since 2005.

Everyone is hopping on DO because it's marketed as a "cloud" and it's part of a tech accelerator (and maybe because it has SSDs and is so damn cheap). Other than that, there's really nothing special about it.

Source: me; hosting-industry expert.

>Everyone is hopping on DO because it's marketed as a "cloud" and it's part of a tech accelerator (and maybe because it has SSDs and is so damn cheap). Other than that, there's really nothing special about it.

The marketing is good. The quality of the product is good. It's incredibly easy to get started with. The tech accelerator I don't think plays any factor. I searched my database and there are 24 mentions of TechStars out of 3213 positive mentions. That's 0.7%, I can't imagine it has much impact right now on their growth. To give some other comps cheap is 115, affordable is 36, ssd is 133, fast is 174, amazing 212, easy is 134.

Source: http://reviewsignal.com/webhosting/company/101/digitalocean

The marketing is false though.

That would mean I can go ahead and market a Galaxy S6 as an Apple device and watch all the fanboys and teenage girls hop on. But of course I need pretty colors.

I measure what people are actually saying about them. People seem pretty happy. You can pontificate all you want, but I am looking at thousands of people's opinions on them, the consensus is that it's pretty good.

If your complaint is the marketing about 'cloud' is false, then cloud computing in general is a pretty ambiguous term. It's being used for things as simple as a VPS in the industry. It's pretty meaningless at this point, people will use it to market their products and I don't think they are wrong. I still couldn't tell you where the line is drawn and I've been around this business for ages.

The biggest reason DO took off has little to do with the reasons you cite. It's got to do with simplicity and hourly billing. AWS had the latter without former. Your typical VPS host has a bit of both but wont tell you launch 20 instances for a few hours for pennies.

> Your typical VPS host has a bit of both but wont tell you to launch 20 instances for a few hours for pennies.

DigitalOcean limits you to 5 off the bat, but that's just being picky.

RackSpace, Atlantic, Gigenet, VPS.net, Oxeo, HPCloud, Upcloud, Dotblock, StormOnDemand (LiquidWeb) to name a few. And that's off the top of my head.

I think the SSDs and the price are why people are jumping on. Not the marketing. Their marketing is horrid.

"You've been developing like a BEAST and your app is ready to go.."

I really loved DO, that is until they emailed me that my server had to be rebooted and I found it with all the config files being replaced by a bunch of @@@'s (I had to use the browser VNC to log in).

I asked them what had happened and if it can be fixed, they tell me to make a new droplet. I ask what will prevent this from happening again... no response...

Bye bye DO.

DO really is a tremendous product. The simplicity, pricing, and speed of setting up an instance is awesome. For setting up a quick dev environment or a simple host for X or Y, it is definitely easier than through AWS.

I use DO pretty often and I hope they have some cool new offerings in the future. Maybe really simple load balancing?

I wish they could address these issues: http://www.redbottledesign.com/blog/five-reasons-why-you-sho...

I think "There's no way to increase storage space without major down-time" is the most important one. One hour of downtime is no good for production.

For what it's worth, #4 is a moot point: DO has said that they're not currently billing for bandwidth overages.[1]

[1] https://www.digitalocean.com/community/questions/where-can-i... (Kamal Nasser February 13, 2014)

For #2 vote for this https://digitalocean.uservoice.com/forums/136585-digitalocea...

and

https://digitalocean.uservoice.com/forums/136585-digitalocea...

I don't think the time behind the scenes will ever be zero so I'd rather 100% be guaranteed not to lose my IP address. Spin up a new VM and assign prod-new the IP of prod-old similar to how you might do it with AWS Elastic IP

I do agree with this. Adding storage to an existing droplet would be nice. Even if it required adding a separate "disk", this would be useful. But obviously, adding, for example, 20GB disk after disk after disk could lead to a lot of annoyance.

You don't care much about downtime if you're using only one droplet for your app in production, do you?

What if you start out small with one server, but then need to expand to multiple droplets? You won't be able to without powering off and making an image.

That's bad planning, but unfortunately also a very common situation. If you really can't afford a couple late hours downtime when you're small then this is what I think I'd do:

- Get a new pair of droplets in front of your current droplet to load balance the traffic (HAProxy, Nginx, you name it). Point the NS records to those :)

+ Why two? Well, you can't afford downtime so there we go.

- Get a new droplet and manually clone the initial one, i.e. copy over config and assets. Add this second droplet to the LB pool for your site/app.

+ Manual work? Well, again you can't afford downtime so you'll need to put some man-hours on this.

- Now you can shut down the initial droplet and image it while the second one gets the traffic coming from the LB.

Eventually you'd want to separate functionality on different servers, so try to also plan that in advance. Also bear in mind that each 512MB costs like $5/mo, that makes this solution work for $20/mo, how cool is that?

I agree there's room for improvement on DO's service, but there's nothing that a good architect can't work around.

You're absolutely right. How about your MySQL write server? You typically have one master that you update to and one or more slaves where you read from. If you get a sudden burst of traffic and your master can't handle all the writes, what do you do?

I understand how HTTP works so I know how to scale it, but unfortunately I don't have the knowledge to scale DB infrastructure.

Anything I say would be a shot in the dark, but I'm interested on your point of view about this one. Caching? Write on two servers? In the second case do you setup a cluster or you just load balance it with a third device/service?

I'm still learning this myself, but from the research I've done I've figured this out:

1. Most people don't go with two writes to a relational database (like MySQL). Syncing is hard, and when it is done, it loses its relational features. 2. Common set up is one mysql write server, scaling it up to fit your traffic, and automatically syncing it with additional servers where SELECT statements only will be made. A random read server is selected if there are multiple to choose from. 3. This also means that in the application, doing a write and immediate read on what you just wrote won't work, so you have to code your application around that. 4. In you app you separate your connections for the reads and writes.

Now, in terms of making changes to the write server is what I'm trying to figure out. Perhaps there's a way to make upgrades on one of the slaves (the reads), then automatically make that slave the master and the master a slave? Not sure, I'm reading a MySQL book (The MySQL Bible) right now that I hope will point me in the right direction.

Well, that was quite interesting, thank you!

No problem!

I wonder at what scale a business like this is sustainable with this low pricing schemas. Whatsoever, will it ever be?

"rejection" -- I had to reread that headline 3 times before I figured out what it was trying to say.

"Reject" is fine, it means "a person or thing dismissed as inadequate or unacceptable."

What the HTML is that?!

...was their catchy phrase on their earlier video commercial on Youtube.

https://www.youtube.com/watch?v=vHZLCahai4Q

I'm sure this contributed in a way to their current success

i dont know why but i'm so tired of these posts

Reply on news.ycombinator.com