I really like the content marketing approach DO took.
If you google for some standard server maintenance problem (e.g. migrating from Apache to nginx), it's usually DO which is ranked first. Their posts are relatively high quality, and there are a lot of them, enough to cover many common problems. For example: https://www.digitalocean.com/community/tutorials/how-to-migr...
It doesn't show annoying pop ups, anything like that, but after visiting those pages many times you might decide to give DO a go. Well, that's what I did.
Also, some people say that similar cheap VPS providers exist for a long time and what DO did was put a nice simple UI. That proves that to create a successful product, you don't need some new complicated features that your competitors do not offer. Making interactions with the system easier, improving user onboarding process by reducing friction can go a long way.
Yep, this is how marketing is done these days. I'm glad someone on HN is appreciating good marketing. I often see devs refer to marketing as snake oil, simply because they've been exposed to bad marketing. Marketing is just like programming, it can be good or bad.
Thanks! Etel wrote many of the articles herself before building out an amazing team. She works really hard to make the articles engaging, informative, and easy to read.
This works super well. It's also the strategy my company employs. Teaching people is the best kind of marketing you can do in the developer space (and probably most any other space).
I think it was Slicehost that pioneered this approach for VPS hosting companies quite a while ago, although it may also have been Linode as I really didn't become aware of them until Rackspace bought Slicehost and I had to migrate.
Similar story here - signed up for DO after coming across a couple of really helpful articles.
For me, the articles show
1) They're competent. If they're experts at solving all these random problems in managing a Linux server, they probably also know how to run a reliable, high-performant VPS.
2) They care about their customers. They're going the extra mile to help customers (and potential customers) solve their actual day-to-day problems. This is a good sign.
Haven't yet had an actual problem that required me to contact their support (which hints at #1), but it's the impression I (and probably many others) got from the articles.
I'm just glad they actually took that idea from them and anybody that provide any services should have these tutorials... It makes it much more approachable for people with little system admin skills.
Loan funding is still funding. In the startup world people think funding== equity funding, but that is not true. Equity funding is normally the most expensive kind of funding. Glad to see they can grow with loans instead.
Actually, in simple terms it's like having a big credit card with $50m in a line of credit. Saying this is funding is like saying I just received $5,000 every time I open a credit card.
Sort of. It's a "line of credit" which, ideally, means potentially money now in exchange for potential future money, more like an overdraft than a loan.
That is, they shouldn't be paying interest on $50m sitting in their bank account right now, but would instead pay interest on what of the $50m they actually choose to use. I don't know if it's different in the US though.
I recall Richard Branson mentioning this sort of arrangement a lot in his autobiography as to how Virgin managed to keep going.
Yes, but not all companies can raise $50mm on credit. Digital Ocean is already well-established. If you're a start-up, people aren't generally interested in investing X dollars at a small percentage return.
Yes, exactly. But getting a loan or a line of credit in the multi-million dollar range implies a certain level of trust in the viability of your company which typically is lacking in most early stage startups. The risk associated with venture capital transforms that relationship into not merely a financial transaction but more of a partnership, which is why equity is typically given.
@edwinyzh as I understand it, basically you deduct the debt from the profit and pay taxes only for the remaining amount. Wikipedia has a very short article on this [1].
Debt itself (as a liability) is definitely not deducted from profit, but repayment of debt is deducted from the profit, and - what is more important - interest on debt is deducted from the tax base as a result. Which means that if you can load up on debt to get higher return, than you would get investing your own money.
It blows my mind that this is happening. Back in the day, there were a thousand cheapo VPS providers out there with similar rates. But DO were the ones to put up an extremely simplified interface to spinning up VMs and that has seen them skyrocket (I'm not saying this is a bad thing).
At the end of the day, I feel sorry for linode. The guys who offer a much better service (in almost every way) and are often overlooked by many customers.
I signed up for DO because they were $5/mo and at the time Linode was $20. Twenty bucks is fine for production but for hobbyist mucking about, which I would classify my usage as, it's a bit steep. I image many users are in a similar situation.
But you're sharing the 1Gb/s link with who knows how many $5 droplets - resulting in unpredictable IO. Linode network IO is dedicated and scalable to 10 Gb/s. DO is shared and capped at 1 Gb/s - even on the largest droplet.
Thanks for bringing this up, this is part of why we closed such a large debt round so that we could really focus on the next evolution of the hardware that we use. There are a lot of things in the works for that and we will be really excited to share them.
Absolutely. I don't need $10 droplets for all of my internal servers, so the cost adds up significantly. I run a cluster of $5 droplets as well as a few large units, and I'm sure many people have similar use cases.
Linode seems to be going strong. I wouldn't dare put anything in production on a rickety shop like DO. Especially after the whole controversy of them taking down someone's blog because it criticized some VP's friend at Google. DO feels like a company founded by devs first and sysadmins last.
DO seems to be for devs farting around and kids putting up wordpress sites. Its not really competing with shops like Linode or Rackspace. That may change, but everything about DO screams, "Do not use me for production."
Meanwhile, my 7 or so production VMs on Linode are going strong and now have been migrated to their new SSD-based platform.
How many startups have devs and no sysadmins? Seems like if a 2 man shop can make DO work without a sysadmin, they just saved a lot of money. When you get big then you can worry about such things.
DigitalOcean offered SSD-backed VMs at low prices. They were one of the first to do so.
I was very happy with Linode, but they dragged their feet for years on it. My workload was I/O-bound and I got far more bang for my buck from DigitalOcean, so I jumped.
By the time DO started competing with Linode, it wasn't worth Linode's time to care about a cheap $5 box. The only problem is that people go to a $5 droplet and stay for a $XXX droplet.
It is so common there is even a term for it, 'Disruptive Innovation' [1].
The term is used in business and technology literature to describe innovations that improve a product or service in ways that the market does not expect, typically first by designing for a different set of consumers in a new market and later by lowering prices in the existing market.
Same here. I still have a $20 Linode for production, but wanting play around and test things out without thinking too much about cost is the reason I also frequently spin up 1-2 DO VPS's.
Linode bet the wrong way by prioritising more cores over SSDs and I think this slip up is what gave DO room to enter the market. Had Linode moved to SSDs earlier and offered their cheapest VM @ $10 instead of $20, DO's trajectory may have been a little different. Also, Linode was (and still is) very bad at marketing imho. DO definitely spanked them on that front (both through their content marketing strategy and just by virtue of their $5 SSD offering which caused all the geek hobbyists to jump over and start blogging about them).
Having said all of that, Linode are still going very strong and remain the much better option for production stuff. I have 3 machines with DO (two $5 ones and a $40 one) and about 15 with Linode (of varying sizes) and in my opinion Linode is head and shoulders above DO in terms of features, capability, professionalism and reliability. They just need to improve their marketing.
Thanks for the feedback, we're definitely looking to improve a lot of the missing features and capabilities and really looking forward to 2015 in that regard.
As for the professionalism and reliability I'd love to hear more as those are items we need to be addressing immediately and the feedback would be very welcome.
Only just seen your reply now. I have no doubt you guys are going to keep on getting better and better and if I were Linode I'd be very worried. I should have mentioned too that it's natural for their to be a capability gap at the moment - Linode's been going since 2003 IIRC so they have had a rather large head start.
In terms of reliability and professionalism, my complaints are very similar to pangram's. Another issue that bit me last week was OpenVPN failing after an apt-get update due to DO forcing a particular kernel outside of the VM.
All of these things are teething issues though and I'm sure you'll get them licked sooner or later (hence why I'm keeping the 3 VMs I have with you).
We switched from DO to Linode a few months ago. I think DO is great, but three things convinced us to move over to Linode. 1) lack of robust logging, 2) we'd get intermittent connectivity problems about once a month (this was in the SF colo), and 3) backups were unreliable, and support didn't really have a good answer as for why the backups were so unreliable. You can send me an HN message if you want more details.
They've been hit with a coldfusion 0day in the past, but the stuff that has happened to them could happen to anyone. I wouldn't have much to critique about their security practices.
That's incorrect along a couple different axes. The biggest problem is, most providers wouldn't be hosting ColdFusion code at all; ColdFusion is archaic and has a terrible security track record. But even if they'd built that UI in Scala, the onus is on the hosting provider to make sure the code is safe.
Being a hosting provider is a big, big deal. They have less margin for security errors than almost any kind of tech company.
Why is Linode so much better? I got a bunch of free credit for DO and I'm only using it for small things so I went with them, thinking about spinning up larger projects however and would like to look at my options.
They've been around a lot longer and have a lot of expertise. I think you'll find that they'll perform better during benchmarking of hardware and have superior network performance and uptime (I've personally experienced times where the entire DO network dropped offline for minute(s)). Their support is great too. I recommend reading up on a few comparison blogs if you wish; just make sure they were written recently and are comparing apples with apples.
But I think the general consensus amongst a lot of people is that DO are better suited to devs wanting to play around and Linode are better for when you want something more than that or want to push something to prod.
edit: now that I think of it, DO developed a reputation for terminating VMs with no notice if you violate their terms or do other things. I personally had my VMs terminated because I signed up for 2 accounts with the same IP. Imagine if I was hosting production content on there....
Their network is more reliable, their control panel is much more advanced and they offer a lot more capabilities (load balancers, longview, additional IPv4 addresses etc.).
DO is great at what it does (I have 3 vms with them) but when you need more, Linode wins.
Last I looked, DO has next to no SLA and Linode offers 99% SLA (almost no cloud provider offers SLA, much less 99% -- exception being Joyent at 100%). Also DO support is known for being a bit slow and Linode support is very quick to respond any time of day. YMMV.
The only thing I'm aware of that, supposedly, made Linode better, is they guaranteed full compute power. If it said you got 1 CPU core, then that was yours, period. From what I understood with DO, their approach was that they'd make every effort to deliver the equivalent compute power of 1 CPU, but would not peg it to you exclusively. I also think this may have changed for DO.
Where have you seen that you have a dedicated CPU core on Linode?
In the VPS market it's nothing unusual to oversell CPU, you can't only have 24 VM on a single server, that doesn't work financially (except if they count threads from hyper-threading as a core but that's not dedicated).
I've been following my DO allocation using New Relic. When I first signed up, I noticed the hypervisor stealing from my allocation. Not a lot, but some. After a week or two, however, this ceased. I can now report that for almost a year now I have been receiving my full allocation. I am very pleased with DO, but based on the good things that I've read on here, I'll give Linode a shot if DO ever drops the ball horrendously.
I'm not sure of the best way, but I was able to tell by looking at the CPU usage chart when you are analyzing your server. The brown line indicates resources stolen by the hypervisor.
> The guys who offer a much better service (in almost every way)
If you need a few small VPSs to play around and test things out, and are fine with the (very) occasional performance / bandwidth degradation, I don't see how Linode offers you a "much better service", at least in terms of cost.
I have a Linode VPS for production, but find myself often spinning up and tearing down DO VPSs for experimentation.
Digitalocean is $5 and Linode $10 for the smallest plans. Linode used to be $20 for smallest, but upgraded their hardware and dropped their prices a while ago.
SliceHost was Linode's only real competitor at the time, they always cost more and gave you less. Rackspace bought them and the pricing pretty much stayed the same.
DO is cheaper but it's not quite the same level of product.
I used several of those 'cheapo VPS providers' in the past, and was always plagued with issues. DO was backed by some major VCs, which made it a bit more trustable.
Recently one of my production node landed on a bad physical host, DO reboot my node twice within 24 hours, I tweeted about it and DO support open a ticket for me.
Since the physical node have problem, I plan to take snapshot then spin up a new node using the image. However, as the physical node have problem, the snapshot take more than 7 hours and still failed to snapshot (No way to cancel snapshot task).
DO support told me the snapshot is running, but I seriously think their support not honest about it and I ask they provide proof to show the snapshot is moving, they failed to do so. (From UI I can see JSON response show progress stuck at 3%, they lied)
Yesterday, DO reboot again my node.
Just my two cent, never ever use DO for serious production usage.
Alternative perspective - Recognize any time you use a cloud provider, you need to be aware that not only are the nodes ephemeral, but the zone/data center you are deploying on may go down as well.
Plan for your server to go down at any time, and, if you have actual $$$ in play, also plan for the zone to go down.
Amazon encourages this thinking by having very low-latency links between AZ (Availability Zones) - and they also recommend deploying in two geographical regions should a regional disaster occur.
I know that all. The point is, DO support lied. After reboot, they still put my node in bad physical host. Also, DO reboot without notice (AWS have EC2 events).
Same happened to one of my production nodes as well. I had to take a snapshot and migrate to another physical node as suggested by the support team. Had about 6 reboots without warning in the last 2 months.
I love DO, but sometimes the reboots without warning can be annoying.
Unfortunately, VM storage isn't a solved problem at scale. You can use remote storage (like EBS) to avoid being dependent on any particular server, but in practice it also has outages and performance anomalies.
it works fine for production. although one time one of my none essential droplet had trojan and started launchign DDOS attacks and it was shut down. I don't know how it happened but the password was super weak.
Strong passwords and always use a private key when SSHing.
I use and enjoy DigitalOcean. Good community docs, good support all around. I've referred a few people to it as well. I'm hoping they keep their support/pricing even if they get big.
OVH offers KVM virtual servers with SSD starting at $2.5/mo[1], and announced they will have an even smaller $1.25/mo plan[2]. All that with more memory than DO.
I'm guessing that VPS providers have a low support cost/customer compared to shared hosting. No hairy compatibility issues, everything is neatly abstracted away, everyone has their own, independent environment with most of the administrative tasks offloaded onto the user.
After you toss in a few guides that double as marketing, and are usually written by the users themselves, all that's left is to take care of the hardware and bill.
Is there anything inherently different (security, performance, or otherwise) about these servers of Runabove's?
I run a fair amount of dev servers, and that price ($2.50) is mind blowing. But i fear that it's far too cheap to be anything but a letdown.. Thoughts?
They're based on OpenStack where DO rolled out their own system, I believe. On the upside, you can use all the CLI tools (nova, etc) that come with OpenStack, existing software should be compatible, and you get some of the additional cloud features (object storage, security groups). On the downside, the panel isn't as nice. The "expert mode" is just OpenStack's Horizon.
But the servers, the instances themselves, are not really any different. I benchmarked the cheapest one for ServerBear[1] a few days ago.
The one thing I miss is IPv6. DO is ahead here.
Would it be OK if I plugged my referral link[2]? With it, you get $10 to test things out.
I was going to sign up, but halfway through realized I can't do it without actually providing a payment method (in spite of the credit). I hope you got the $10 credit for my initial sign-up though.
I think where they win is when somebody buys a $5/mo server and uses $0.50/mo worth of capacity. They shut down all the hardware they don't need and save the electricity costs.
We're growing so fast that shutting down hardware is the last thing we would ever do. We have the opposite problem -- it's quite challenging to provision hardware fast enough to keep up with demand. :)
Haha, that's surprising. I've heard lots of stories about load balancing VMs to certain corners of the datacenter so the rest can be shut down that I figured everybody had unused metal at least at some point during the day.
They're hardly the only VPS provider offering that price point. (Although it's true, many of the others have a slightly fly-by-night look about them. But by no means all of them.)
Sadly 99.9% of the other companies are shitty solusvm customers who dont understand the hypervisors they are using, and completely fail to manage resources properly, resulting in retarded TOS such "a load average of more than 1.0 for more than 15 minutes" results in warning/suspension/ban. You'd be insane to trust any production stuff to retard kiddie hosts from LowEndBox or WHT.
Definitely. These are good for VPN/proxy endpoints in various geographic areas to test servers or get around restrictions and not much else. I use the cheap VPS boxes for backup static file hosting for my products, but I serve everything primarily off Amazon S3.
I never had a problem paying via debit card when I used them. I don't have any active droplets with them right now, but I've always paid with my bank issued debit card, and looking at it now it's still the active payment method on my account.
They also don't accept prepaid cards since they've said 80% of fraud cases involve prepaid cards. Linode took my money and worked well but they don't have a $5 price point.
If you google for some standard server maintenance problem (e.g. migrating from Apache to nginx), it's usually DO which is ranked first. Their posts are relatively high quality, and there are a lot of them, enough to cover many common problems. For example: https://www.digitalocean.com/community/tutorials/how-to-migr...
It doesn't show annoying pop ups, anything like that, but after visiting those pages many times you might decide to give DO a go. Well, that's what I did.
Also, some people say that similar cheap VPS providers exist for a long time and what DO did was put a nice simple UI. That proves that to create a successful product, you don't need some new complicated features that your competitors do not offer. Making interactions with the system easier, improving user onboarding process by reducing friction can go a long way.