5 ms·
Ask HN: Best VPS providers for fast scale-out
I am searching for a good , cost effective VPS provider. My main decision point will be how quickly I can spin up new instances (minutes is no good I need seconds). Any recommendations? THANKS!
- benologist 14y agoHeroku can spin up new instances in seconds and you can automate it with HireFireApp. It's not a VPS, but a VPS aka shared hardware is never going to be "seconds" unless it's already provisioned and basically just needs to boot (EC2 and your own AMIs?).
- ngcoders 14y agoLinode is good with nice API , Even Digital Ocean is good with basic API. We are using DO ( value for money ) for a new service and A full nginx based site can be setup in under 60 - 80 seconds after cloning a basic image. New instances take a little less time.
- platypii 14y agoFriend of doppenhe here. Amazon, heroku etc aren't fast enough for us due to spin-up time. Basically when we get a particular command from a user, we need to do a LOT computation on many parallel machines, but for a short burst of time. Keeping a whole cluster of machines reserved would be prohibitively expensive at this stage. We would prefer to keep a small base-load capacity reserved, and then spin up a large number of instance for a short time. Response time is key here. It would be best if they were already up and running, waiting for a command, but idle most of the time. Fine if it's shared with others. Kind of like a computation insurance for short bursts of high-demand. Anyone know any services like that?
- benologist 14y agoWhat about cheap dedicateds? In the US Hivelocity have great prices - http://hivelocity.net/ http://hivelocity.net/ In Europe there's also Hetzner - http://www.hetzner.de/en/ http://www.hetzner.de/en/
- kintamanimatt 14y agoThere's always going to be some time associated with creating a VPS. You're not going to get a new VPS instance allocated in seconds simply due to these practical issues. For this reason you may need to keep some idle servers running if you really need a high level of responsiveness. It sounds like your real issue is cost. Perhaps two options you might consider: buy (don't rent) your own hardware and get some colo space nearby. It can work out significantly cheaper and neither the hardware or rackspace has to be expensive. The second option to minimize costs would be to rewrite the key parts of your algorithms in something like C, if you haven't already. If you're doing the heavy lifting in something like Ruby, or even Java, you're going to be wasting a lot of resources in a situation where it sounds like every cycle counts. Hope this helps.
- tectonic 14y agoAmazon EC2 Spot Instances?
- sheraz 14y agoI recommend you take this question over to the WebHostingTalk forums [1]. They have a lot of smart people who are in that space and can offer a lot of insights as well as deals on hardware, bandwidth, storage, etc. I am not affiliated, just a user there from time to time. [1] - http://webhostingtalk.com http://webhostingtalk.com
- doppenhe 14y agothanks!
- Pyramids 14y agoPerformance: - PhoenixNap SecuredCloud http://www.phoenixnap.com/ http://www.phoenixnap.com/ (US-West, US-East, Amsterdam) [VMWare based, spinup time is about 45 seconds, offers every imaginable feature for large production environments, enterprise firewall, f5 load balancing, 100% uptime SLA] Value: - Digital Ocean https://www.digitalocean.com/ https://www.digitalocean.com/ (US-East, Amsterdam) [KVM/QEMU based, spinup time is under a minute, decent 'basic' API] As a current customer of both companies, we use Digital Ocean to host our company VPN and development nodes, and PhoenixNAP to host our production sites. (Bandwidth at PhoenixNAP is billed @.05- per gb, very competitive compared to EC2 especially considering they use exclusively premium bandwidth)