New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
The most premium and high uptime VPS for production use
My budget is $5/month (can be 10 depending on situation/specs/reputation), reputable provider only. No ovh, linode, do, anything in quadranet, oplink, buyvm, etc, please don't suggest these.
I need:
- 128MB ram
- <15 GB bw
- 2-3 GB disk (maybe less)
- US West non-Los Angeles
- Good peering/latency to AS64247, AS73
- any real virt is fine.
Not really any other requirements. I am running only a TLS-client-certificate-required non public webserver so I need UDP 443 and UDP 80 (not tcp 80).
Comments
Vultr has US West in Seattle and Silicon Valley.
increase your budget and we can start discussing HA VMs with failover to another DC.
Single VM is fine, this is going into an existing cluster
Both already taken care of, any other providers?
How about SecureDragon? http://drgn.biz/
@virmach special offers.. They have lot of package.. You can choose the available one..
https://virmach.com/special-offers
They only offer openvz in the right location.
Maybe RamNode/Incero in Seattle then.
Also done in speedykvm, and CNservers seattle
Are there any less known hosts or something? Maybe in portland?
UltraVPS - Vegas
NQHost - Phoenix
InceptionHosting - Phoenix
http://www.pdxhosting.net/about/datacenter.php I've never personally used them though.
I doubt if Virmach is capable for being used in a production environment since their servers are kind unstable recently
Perhaps the single core option? https://www.nfoservers.com/order-virtual-dedicated-server.php
Have proofs?
I personally don't mind this kind of issues since I'm not doing critical stuff, but for someone that uses it in a production environment, it might be an issue (at least before the issue get fixed)
Have all of their locations, thanks!
Yeah I absolutely can't have any reboots - the application I'm running requires human intervention to start/restart and enter something in a keyboard interactive session. Suspending VM is fine, but if it actually does a powerdown it's not going to fix itself
They used to be rock solid tho, just have to give them some time.
I highly recommend forking a little bit extra out and trying @clouvider network
Worked wonderfully for me so far in their NY location, like this:
7:15AM up 183 days, 9:17, 1 user, load averages: 0.21, 0.14, 0.09The email @Lavenade quoted was for the Dallas location though...
Webserver on UDP, are you sure you know what you are even talking about.
As for the topic, rebuild your app so it doesn't require highest uptime from any single VPS, and instead utilize redundancy across providers and continents.
Bandwagonhost never let me down. Check them out. They're one of the few I've been using for years.
Yes. All client IoT devices present client certificates for authentication, the servers all present pinned certificates, and transport is basically HTTP/2+TLS-via-UDP; the protocol has been used in production for a very long time for a lot of people: https://www.chromium.org/quic
It's already redundant across providers (i'm looking for more providers and some diversity, not one place to run it), but currently still requires fairly tight timing tolerances and signed responses may be invalid if they arrive >100ms later. After a device registers itself to a load balanced shard it'll stick to it, but DNS TTL and whatever will potentially result in lost data when something fails over / reconnects
Yes, ive seen it after i already posted this. Anyway wanted to say that i am quite satisfied with the quality of virmach vps-es, but yes, i am still cautious about production use because of the price too...
Damn i remember my time with them, had a vps with ipv4 for 4 or 5$ a year and it was never unreachable in the whole year.
I'd probably say DigitalOcean. But I want to stress the only HA is redundancy.
You can have the most reliable company ever but there can always be circumstances out of their control. i.e regional issues.
Yep, my BandwagonHost VM is quite stable too, but OP has clearly stated OVZ is unacceptable.
most premium and high uptime VPS for production use
Budget $5
Perhaps you can use SingleHop, but it's too expensive.
~ 50 USD Per Month
@hzr i know Miami is east but would you run some tests on our looking glass there : http://lg.miami2.ikoula.com/ and tell me if that could do ?
ImpactVPS and HostUS are solid providers
I do apologize for any inconvenience. I would like to go over the situation in this thread for anyone that is interested, as this is an isolated issue.
I do admit this took longer than we would have liked, but this was a very abnormal issue. The datacenter was also involved, which further delayed a resolution. With all that said, I believe it's unfair to state that our servers are unstable, as that makes it sound like all our servers had this problem. In the last two months, an extremely small number of customers were negatively impacted by these problems. This number may be relatively high to our normal downtimes, since multiple identical servers were affected, but still low. A server that was negatively affected would have close to approximately 99.9% uptime, as the problem occurred every few weeks and downtimes were handled immediately by our team.
Even with these negatively affected servers, our average uptime across all servers in the past two month has been approximately 99.99% or more.
Rest assured that our pricing has nothing to do with the quality of our service. We may have some specials with limited customer support, but all nodes are managed extremely well by our talented, growing team.
Since six months ago, we have invested in many improvements at no additional cost to customers. This includes extending support hours during the nights and weekends, providing free backups on most servers, adding a live chat for basic inquiries, providing automatic downtime credits, hiring new support agents with more hours covered overall, extending our credits & refund period, upgrading our websites/portals, and fine-tuning our automated systems.
I do want to let you know that all the nodes mentioned above were immediately locked off since the issue began, so anyone who ordered would be placed on a stable node. However, some downtimes are bound to still occur, even though the likelihood of you specifically being affected within the next year is very low.