Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


What to pick for maximum uptime?
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.

What to pick for maximum uptime?

kalimov622kalimov622 Member
edited May 2019 in General

I currently have an OVH dedi but I want to cut down some of the costs since my app barely uses 2 to 4 GB of RAM and 1 CPU core anyway and I was recently looking for alternatives that would bring me the highest uptime with minimum to no downtime of all if possible. I'm aware this is kinda of a lottery but if I'd get the same uptime I currently have with OVH I'd be happy as my server has never been offline more than a few minutes in the past several years, last time I think it was last year when they had some network issues and thousands of servers went down.
It's a TCP app (not website) that uses a fairly small database of around 300MB as well as about 200K small files (no more than 16-32KB) that I automatically backup every hour or so and I was currently considering the following options in EU:

Kimsufi: KS-9 Intel Xeon W3520 4c/8t 2,66GHz 16Go DDR3 ECC 1333 MHz SoftRaid 2x240Go SSD
OVH: VPS Cloud 1 1 vCore(s) From 3 GHz 2 GB RAM 25 GB
NFO: Two virtual CPU cores (Intel Sandy Bridge or better) 2048 MB of RAM 200 GB of RAID
Any other options? Only interested in good uptime, stable network, fast support in case of hardware failure.

Few things I'm not sure about is how stable is OVH cloud? Is the uptime good, are they rebooting servers often? That would be my main concern in terms of VPS for OVH and NFO together. As for Kimsufi, I have a server there with pretty much 100% uptime that's still in the OVH management panel, never had any hardware issues but last time I checked they removed any contact forms except a ticket/forum where you had to submit some SMART data and only then you'd have your issue looked at and eventually hardware replaced if anything was wrong with it and their panel is now separated from the OVH one as well. I've seen people reporting they had to wait even a couple days/weeks to have their faulty HDD replaced there and that's definetly something I want to avoid as I'm aiming for strong uptime.

Comments

  • FAT32FAT32 Administrator, Deal Compiler Extraordinaire

    Bare dedicated server doesn't really guarantee you the best uptime unless you are doing clustering. I would suggest running it on 3 small VPS on different providers and regions, cloud can be an option too.

    It is quite hard for all 3 to be down at the same time, yet you can sync up between VPSes in case one of them went down.

    Thanked by 2kalimov622 uptime
  • kalimov622kalimov622 Member
    edited May 2019

    Thanks, I probably had to mention that the app is TCP and while syncing database it's not a problem doing the same for small files that are not part of the database it's going to be problematic as they update quite often, besides every user connected to the app will be kicked out whenever the server goes down.

  • poissonpoisson Member

    You need to ask @uptime about this.

    Thanked by 2uptime Janevski
  • uptimeuptime Member
    edited May 2019

    i try not to judge but cannot help but notice many run-on sentences in the OP with no punctuation but a certain cadence which is familiar territory in cases where stimulants may not have been properly titrated and i would suggest at this point that moderation is key to a successful outcome so it may be beneficial to take a nap from time to time possibly try taking some vitamin C (but not grapefruit juice though grapes should be fine if you peel them carefully) so as to accelerate the excretion of accumulated metabolites and so forth and so on regardless while it may seem productive in the moment to push the limits some downtime is actually occasionally necessary to maintain functioning of a healthy system so i do try to take mine offline to update software and reboot on a regular basis otherwise an exponential accumulation of impaired judgement and other generally unpleasant sequelae may become evident as time goes on and on and on and on and on

    EDIT2:

    That said ... I would tend to agree with @FAT32 's suggestion for high availability based on a more distributed system. This may introduce a challenge to maintain consistency across different nodes of your database, but that is a well-studied and generally solvable problem.

    I've been happy with reliability of the cheap kimsufi boxes in both France and Canada but (vaguely) recall hearing mixed reviews about OVH cloud.

    Really, just give it your best shot - test a few options over a period of time - and be prepared to make changes. It's helpful to try to plan ahead but really seldom possible to anticipate every wrinkle. And - in my opinion - probably that much more so if one is basing their plans on advice from random strangers on a forum primarily devoted to the pursuit of cut-rate services.

    (Though it is likely that others here can offer more useful advice.)

    Good luck with your uptime!

    Thanked by 2angstrom kalimov622
  • angstromangstrom Moderator

    @uptime said: Good luck with your uptime!

    Please try to refrain from talking about yourself so often.

    Thanked by 2poisson uptime
  • poissonpoisson Member

    A question about uptime obviously requires @uptime to talk about uptime.

    Thanked by 1angstrom
  • uptimeuptime Member

    and downtime, lest it become upside-downtime ...

    Thanked by 2angstrom poisson
  • @kalimov622 said: 2 to 4 GB of RAM and 1 CPU core anyway and I was recently looking for alternatives that would bring me the highest uptime with minimum to no downtime ... in EU

    For your rather modest requirements (resource wise at least), I'd say go with a good VPS provider and you'll be all set. I can definitely recommend Netcup (if you're lucky you can still pick up an Easter egg or two). A Dedi definitely seems a bit over the top.

    You can't go wrong with @UltraVPS (see this thread) or with @LiteServer (their NVMe KVMs are blazing fast) either.

    I haven't had any downtime on any of these providers (or rather all downtimes are PEBCAK).

  • We need new username for @downtime , so many member hates him

    Thanked by 3uptime ehab coreflux
  • angstromangstrom Moderator

    @ErawanArifNugroho said:
    We need new username for @downtime , so many member hates him

    Please don't give @WSS ideas.

  • donlidonli Member
    edited May 2019

    @kalimov622 said:

    ...

    I was recently looking for alternatives that would bring me the highest uptime with minimum to no downtime of all if possible.

    See: https://landing.google.com/sre/books/

    Thanked by 1uptime
  • YuraYura Member

    Redundancy.

  • uptimeuptime Member

    Redundancy.

    Thanked by 1Yura
  • donlidonli Member

    @uptime said:
    Redundancy.

    You mean:
    Redundancy. Redundancy. Redundancy.

  • Hetzner.

  • YuraYura Member

    @donli said:

    @uptime said:
    Redundancy.

    You mean:
    Redundancy. Redundancy. Redundancy.

    That was redundant.

    Good.

  • uptimeuptime Member
    edited May 2019

    @kalimov622 if OVH server is overkill for your application but their infrastructure has been working well for you, then you might do well to check out some of the VPS products from wishosting - both to reduce costs and to increase opportunity for support in case of hardware failure.

    While I personally haven't been using them long enough to seriously evaluate their long-term uptime, they seem pretty solid and have a good reputation for handling their business without any drama as far as I can tell. They are on this forum as @exception0x876

    I mention them because many of their offers are based on OVH servers. Not particularly high-end systems - but well balanced and reliable to get the job done at a lower price point. Worth taking a closer look in my opinion.

    Seems like you may be looking for more suggestions for providers along these lines?

  • deankdeank Member, Troll

    Redundancy is for pussies. Real men take risks.

    Thanked by 1coreflux
  • YuraYura Member

    @deank said:
    Redundancy is for pussies. Real men take risks.

    I cun't agree enough.

  • Suggestion - you mention OVH Cloud but what about Kimsufi SSD VPS offerings? That and another LE VPS provider for redundancy, as previously suggested.

  • kalimov622kalimov622 Member
    edited May 2019

    @uptime, exactly. OVH has been great so far for me but I have no experience with their VPS range and I'm not sure how stable they are and if they reboot their servers often for example. I've read some of the older posts here and feedback is like 50/50 but I guess things have changed a little these days. For example Hetzner cloud used to be quite unstable at first and so far except couple random reboots they seem to be pretty solid as well.
    I am open for more suggestions as well and while I don't have any experience with providers like NFO, vultr, ultravps they've been around for a while and seen many recommending them.

    @ruggieroav, I was actually considering KS-9 W3520 but since Kimsufi has a separate website/panel now I'm not sure how fast the hardware support would be. If it would take days for them to read those tickets or look at their forums it's clearly not going to work for me.

    Due to the nature of the app it won't be the same as setting up a HA for a website and have multiple vps so the data can be read from the other online servers if one goes offline.

    Thanked by 1uptime
  • YuraYura Member
    edited May 2019

    .

    Thanked by 2emgh Janevski
  • @Yura I agree.

  • uptimeuptime Member

    @kalimov622 - I see.

    as far as suggestions regarding reliable "low end" providers ... you're going to want some way to vet the reliability of the suggestions as well :) I'm just one of many "enthusiadts" here with no real skin in the game. Ultimately it's going to be up to your own judgement and preferences.

    I think there are quite a few decent choices but reliability tends to cost something to provide. Quality hardware, datacenter and network infrastructure, technical expertise, support .... none of these things spring fully formed out of thin air. So generally speaking there are going to be some trade-offs with regard to cost. And even the best are not perfect. (I've found consistent network performance to be the weakest link - that's a tough row to hoe. Superior potassium can cost many bushels of potatoe in that department.)

    It might be helpful to get a better picture of how much you are spending currently, and what your target budget is?

    Anyway, here are a few more good prospects (well-established providers) for decent reliability in Europe that I've tried long enough to recommend considering:

    • AlphaVPS.bg (not to be confused with AlpoRacks)
    • Budgetnode
    • CrownCloud
    • First-Root
    • Hetzner
    • HostHatch
    • Inception Hosting
    • Liteserver
    • UltraVPS.eu (not .com, that's someone else)

    I'm a lightweight user and don't ever have to seriously deal with "urgent" support experiences so can't really speak to that except to say some seem to have a better reputation for timely support than others (but from my point of view they've all been reliable enough that I have never needed to stress about support).

    if you do some research you'll find some complaints about a few of these (Budgetnode for example) so be aware that YMMV but I've been happy with general quality and uptime from all of them so far.

  • SirFoxySirFoxy Member

    @Yura said:
    .

    ..

    Thanked by 1emgh
  • @uptime said:

    • AlphaVPS.bg

    I'm going start using this exact string whenever I mention them.
    Reminds me of ultravps.eu(Bradler&Krantz Gmbh)
    vs
    ultravps.com(Versaweb's brand)

    Thanked by 1uptime
Sign In or Register to comment.