Results 1 to 10 of 10
  1. #1

    should i buy a backup server to prevent downtime?

    I currently have a dedicated server from 1&1 (Windows Server 2003).

    The only problem I'm having is that sometimes I need to restart the server, and when I do, all of the websites I'm currently hosting are down for a minute or two.

    This is unacceptable, and I'm currently looking for a solution.

    I was thinking of renting another (cheaper) server from 1&1 and then load-balancing them using NLB, but I don't know if this is possible.

    Does anyone know if this would be possible to do? Does it sound like the right solution? Are there any other solutions that might work better than this?

    Basically, all I want to be able to do is "switch" the flow of traffic from one server to the other when I have to reboot the machine. If I can load-balance the two servers while I'm at it, then that's fine, but my biggest concern is avoiding that downtime when I have to reboot my server.


  2. #2
    Join Date
    Jan 2004
    North Yorkshire, UK

  3. #3
    Join Date
    Sep 2003
    Unless your customers are paying you in the $xxxx i wouldnt worry about 2 minutes of downtime.
    Centation Web Services
    Bristol based web design
    Offering website design, SEO, website hosting, website development and domain registration.

  4. #4
    Join Date
    Apr 2005
    Restart the server at 3 in the morning. The solution you want probably wont work. Normally it would be best to move the ips from one server to another. This would cause an immediate transfer of the traffic from the main system to the backup. Normally such systems are a cluster. The disk data and everything is replicated from one to another or they shared a disk subsystem, with only one using it at a time. The only other way to do it would be by dns and typically you couldnt guarantee that the dns would change so that the websites would be accessable on the second ip address. I think you are out of luck unless you decided to colocate your own two machines someplace and set them up as a cluster.

  5. #5
    Join Date
    Jul 2005
    As I am new to servers I have the question why do you have to reboot your servers so often? I will have linux, I am not willing to reboot several times a week

  6. #6
    Join Date
    Apr 2005
    He said "sometimes" not "several times a week."

    Maybe hes installing some applications on the system that require a reboot or hes installed a security patch that occassionally requires a reboot. I would say when we install security patches we may have to reboot once per month. (maybe) I guess I would have to start keeping logs to find out how often we reboot.

    EDIT: Please dont start a windows vs. linux argument in this thread. It is very annoying.

  7. #7
    Join Date
    Jan 2001
    Miami, FL
    you will need more then just a load balancer here... when we construct Fault Tolerant clusters for our enterprise clients, we utilize load balancers as well as a centralized NAS.
    Biznesshosting, Inc. DBA VOLICO - Intelligent Hosting Solutions
    East Coast Enterprise Dedicated Servers and Miami Colocation.
    managed and unmanaged dedicated servers. High bandwidth colocation. Managed clusters.

  8. #8
    wow, thanks for all of your comments --- i wasn't expecting so many replies so quickly!

    anyway... yeah, i restart my server once a week at the moment because i'm running a custom-built application which acts as a game server of sorts, and it leaks memory.

    I know that the real solution to this problem is to sort out the dodgy application, but... If I install software on the server which requires a restart, then that means downtime... If I install a security patch, it might require a restart... I know that 2 minutes doesn't sound like a lot of time, but I'm trying to run a professional website, and - as I mentioned above - 2 minutes downtime is unacceptable.

    At work, we use NLB (network load balancing) on our Win2k3 servers, and it works brilliantly. On the odd occasion when we have to restart one of the servers, all we do is "drainstop" the server we need to restart which forces all the traffic to go to the other server, and then as soon as we bring up the other server, they both start accepting requests.

    This is exactly the type of thing I'd like to be able to do, but I just didn't know if I'd be able to do it with 1&1. I've e-mailed them, but I haven't heard anything as yet.

    I was just wondering if there might be a better alternative that I wasn't aware of.

    Thanks again for all your help. I'll keep checking back to let you know it's going and to see what you guys suggest.

  9. #9
    You could do a failover via dns? Get a third-party to do the DNS, like or something to that likes, some of them offer the failover service and you can set it up accordingly so that it goes to your backup.

  10. #10
    It's going to be very expensive, but is possible - though i do not know if at 1&1

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts