Howdy, Stranger!

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


Which Lowend KVM Providers has >99.97% uptime for the last year?
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.

Which Lowend KVM Providers has >99.97% uptime for the last year?

serverianserverian Member
edited July 2013 in General

^

Comments

  • InfinityInfinity Member, Host Rep

    None?

    I'm pretty sure that means less than an hour of downtime in a whole year unless my math is completely off.

  • Yup 99.99% uptime is only 52.56 minutes for a year
    http://en.wikipedia.org/wiki/High_availability

    If you want that much uptime @serverian get a dedi

  • @Infinity, yes 52.56 minutes.

  • Awmusic12635Awmusic12635 Member, Host Rep

    Well, 100% for the last month.... only have really offered it that long.

  • Okay, let's say 99.97% then, 2 hours of downtime in a year.

  • InfinityInfinity Member, Host Rep
    edited July 2013

    Well I have had pretty much that much uptime with Go-VPS-Go although I don't think they do low-end anymore, they had a nice deal when I picked it up. I can't say I've had as much luck with any other provider, there are always problems here and there especially at this price point.

    Having said that Castlegem, although I haven't been for a full year have pretty much been up for the whole of the period I've used them (maybe 2/3 of a year) with about 30 minutes downtime at most.

    But obviously there are different nodes and can be different problems etc. ;)

    Thanked by 1ozfingwe
  • Got no downtime so far with few nodes @ vpscolo (Sis.Solarvps) on Peer1 DC, controlling since 3years for clients. Not KVM, also no stock since last 1 year :D

  • Why? Thinking of breaking the record, @serverian?

  • jhjh Member

    99.98% :)

    Server's full and not getting another one though.

  • mpkossenmpkossen Member
    edited July 2013

    My Prometeus servers haven't had any downtime at all, for almost a year. Except for my own reboots for kernel updates, that is.

    Thanked by 1Infinity
  • SimpleNodeSimpleNode Member
    edited July 2013

    We've had about 99.99% uptime over the year that we've been running.

    Our KVM plans are here: https://simplenode.co/kvm

    Status page: https://simplenode.co/status/

    ServerBear Uptime:

  • In Pingdom (1m resolution) probes for each server are either 100% or 99.99%.

    ServerBear says 100%:
    image

  • Based on my experience:

    Hostigation and prometeus so far are the providers with the best track in uptime.

    BuyVM has had a couple of issues on their datacenter (electrical problems at Nevada).

    TinyVM (ramhost) the issues they had were due to their datacenter (networking related)

    Ramnode has a good uptime history but they dont have the year youre asking for.

  • Downtime with or without maintenance notification? Only with ksplice or something it is possible because you must upgrade your kernel and update the software..There's more to uptime than that..

  • MaouniqueMaounique Host Rep, Veteran
    edited July 2013

    We have had no reboots on KVM since the servers were put up:

    [root@pm19 ~]# uptime
    03:04:54 up 407 days, 13:18, 2 users, load average: 5.44, 6.01, 6.19

    That is the first iirc, 24 threads dual e5.

    A few VPSes with ISO still mounted crashed when we switched storage to NFS.
    However, due to kernel problems, everyone having multiple users on the kvm should have updated theirs.

    Since KVM nodes do not allow access to the host kernel to any other user than us, the admins, we didnt need to update it.

    Also, Xen:

    [root@pm12 ~]# uptime
    03:09:26 up 255 days, 5:26, 1 user, load average: 0.00, 0.05, 0.04

    Doesn't have an year yet, Uncle started offering xen later. These are centos 5 not affected by the kernel previous exploit and also not by the Xen bug (special crafted kernel) since that kernel cant be loaded on them.

    However, we do have issues with OVZ. Not only the privileges escalation bug, but also the DoS problems plagues us so much so that we dont even schedule maintenance to update kernel, we update it and will be loaded next time it crashes, if it is a race condition, or DoS types of exploit... (well, except the privilege escalation one, we are jalous and dont want more superusers one the nodes :P )

    Still, most have uptime like this:

    [root@pm50ovz ~]# uptime
    03:14:59 up 51 days, 22:01, 2 users, load average: 2.03, 2.05, 2.15

    Since the kernel patching, that is.

    However, this is only part of the story, we had many DDoSes lasting for a few minutes with a lot of packet loss (lot less on KVMs or Xen, though, most on OVZ nodes) as well as one IPv6 outage for a few hours a few core router reboots and switches replacements, so, it depends on how you consider uptime as an uptime of services or the node.

  • it depends on how you consider uptime as an uptime of services or the node.

    From an end user perspective "availability" might be a better word to use than "uptime". If a node has 200 days of uptime but the network has frequent outages or packet loss that makes the VPS unreachable then that 200 days uptime is meaningless.

  • Linode 512 on UK

    root@s1 [~]# uptime
    00:27:23 up 632 days, 14:07, 1 user, load average: 0.04, 0.03, 0.05

  • bcrlsnbcrlsn Member

    On my dedi even with switching over to a new server my main mirror has an uptime of 99.76% in the past 6 months.

    I had a bunch of small (2-4 minute) network outages over at WSI. New DC and it's nearly 100% up!

  • asterisk14asterisk14 Member
    edited July 2013

    @DomainBop said:
    From an end user perspective "availability" might be a better word to use than "uptime". If a node has 200 days of uptime but the network has frequent outages or packet loss that makes the VPS unreachable then that 200 days uptime is meaningless.

    ^^^ This!

    Maybe something like availability information should be included in the offer LEB providers make. I think some of the providers like @BlueVM have status pages which is a step in the right direction. Like I said in another thread people want a RELIABLE service, not necessarily the best specs. I'd rather have a Toyota than an Alfa Romeo!

  • EhtiEhti Member
    edited July 2013

    My RamNode VPS had pretty good uptime before they went through that hack, but don't remember the exact figure sorry. Still, it's been running solid after the restoration with no downtime!
    EDIT: Err sorry mine's VZ, not KVM.

  • 99.4%. Thats <99,7%, but not a lot. At VPSCorner, according to geekahost. http://geekahost.com/db/show.php?mode=overview&server_id=32#.UdiKfBbolyY

Sign In or Register to comment.