Slashdot Mirror


OpenBSD Acquires IP Load Balancing

xarc writes "OpenBSD 3.2-current has acquired IP load balancing support via its packet filter, PF. This is a great step for those of us who prefer OpenBSD, but are dependent on other OSes and software (such as Linux's Linux Virtual Server) to provide similar functionality."

8 of 19 comments (clear)

  1. uhm by Anonymous Coward · · Score: 1, Interesting

    did you even read the link you posted? this is a far cry from load balancing.

    1. Re:uhm by sporty · · Score: 3, Informative

      Far cry nothing. Load balancers do use the roundrobin and hash algorithms.

      --

      -
      ping -f 255.255.255.255 # if only

    2. Re:uhm by psyconaut · · Score: 3, Insightful

      True, load balancers do use round-robin and hash algorithms....but these days a load balancer also implies some logic to stop it forwarding requests to a none-active device. i.e. some sort of keep-alive mechanism.

      But, to see this is OpenBSD is nice. I might actually consider writing some hooks myself to make it a bit more usable.

      -psy

    3. Re:uhm by Anonymous Coward · · Score: 1, Insightful

      This is what userland script should do ... This should not be inside the kernel.

    4. Re:uhm by evil_pb · · Score: 2, Insightful
      And it's also just a first update into the CVS tree, mid-release, which already does MOST of what a Load Balancer would do! It may not detect down hosts (yet), but given the usual quality and completeness of OpenBSD code I suspect by 3.3 this will be ready for prime time. I'm not sure what else you think it's missing - load balancers are not highly complex pieces of equipment.

      And this isn't just an IPfilter theft like someone else had stated. PF is a complete re-write, with dramatic performance gains and cleaner integration into the system. (By gains, I'm talking about documented increases by orders of magnitude, with large rulebases.) I suspect the LB code is no different, probably all new code. It'e been submitted by Daniel Hartmier as well, who wrote PF, so I'm fairly confident that the code won't suck.

      Now I just want to see failover capability in PF. I know the VRRP licensing BS is what is stopping that effort right now, but I hope that can be solved soon (a new protocol needs to be created IMO). The only reason OpenBSD can't be deployed in many areas is the lack of failover, that's such a critical piece of infrastructure these days that it's necessary.

  2. wait for it by JamesCronus · · Score: 2, Funny

    hang on, where's the obligitory " BSD is dead" post?, if its dead, how come stuff like this is getting released, it looks like bsd is focusing on its niche market, servers, i'd say load balancing is quite important there..

    --
    dybia felly dwi a hampster (i think therefore i am a hampster)
  3. Re:ipfilter anyone? by Anonymous Coward · · Score: 2, Insightful

    They didn't copy the code from ipfilter. It supports much more than just plain round-robin.

    Addresses can be allocated in a number of ways:
    - masking out the network portion of the address and replacing it
    - randomly assigning an address in the block
    - hashing the source address and a key to determine the redirection address
    - iterating through the addresses sequentially (this is the only allocation scheme which works when a list of addresses is specified)

    It also supports load balancing on route-to, dup-to and reply-to. Again something that ipfilter doesn't.

  4. A well thought out move by PFAK · · Score: 1

    This is good on Theo's & OpenBSD's part to have load balancing in the codebase, this should fill a niche in the server market, and maybe catch OpenBSD up to Linux a tad more.

    Give them credit, they have one of the most mature and stable OS'es out there. And Theo is sure smart.

    Oh, yeah and the best thing. It's Canadian ;)

    --

    Free means no restrictions, ironic the FSF's GPL forces restrictions, isn't it? What's your definition of free?