You are viewing an old version of this page. View the current version.

    Compare with Current View Page History

    « Previous Version 5 Next »

    There are several way to control which hosts the supervisor will select when deciding where to start job instances on the farm.


     

    Job and Worker Restrictions

    Restrictions are used to allow or restrict where jobs run, and are applied to both jobs and Workers.  Restrictions are based on cluster names. A job has preferential priority on a Worker whose cluster matches the job's cluster, but the job is free to run on any Worker in any other cluster, subject to the restrictions defined here.

    Restrictions defined for jobs 

    When a job has a restriction defined, it means only run on hosts that satisfy the restriction expression. Hosts that don't satisfy the restriction expression won't be considered as dispatch candidates (the job will never be sent to that Worker).

    Restrictions defined for Workers

    When a Worker has a restriction defined via its worker_restrictions value, it means only run jobs whose cluster value matches one of the clusters in that worker's restriction expression.  The worker won't accept jobs whose cluster doesn't match one of the clusters in the worker's restriction expression.

    Restrictions Syntax

    A restriction is really defined as a "filter" for hosts based upon information in the queuing algorithm; the values are one or more cluster names. In the priority/cluster queuing system, a user specifies the restrictions with a directory structure format:

    /[<segment>/][<segment>/][+|*]

    • * means only the first level below.
    • + means all levels below that level, regardless of depth in the hierarchy.
    Icon

    The restriction value is actually evaluated as an expression, and multiple clusters are specified in a "this cluster OR that cluster OR the other cluster" type of string, with the " || " symbol to mean OR.

    Examples

    Job Restrictions

    SyntaxMeaning
    qbsub -cl /private -restr /private <cmd>Submit a job that will have highest priority in /private and run only in /private
    qbsub -cl /private/very -restr '/private || /private/*' <cmd>

    Submit a job that will have highest priority in /private/very, but could run in any host in
    /private or in the first level below /private

    qbsub -cl /private/very/deep -restr '/private || /private/+' <cmd>

    Submit a job that will have highest priority in /private/very/deep, but could run in any
    host at any level at /private or below

    Worker Restrictions

    SyntaxMeaning
    worker_cluster = "/private/very/deep"
    worker_restrictions = "/private/very/deep"
    Define a host that will only run jobs in /private/very/deep
    worker_cluster = "/private"
    worker_restrictions = "/private || /private/*"
    Define a host that will run jobs in any cluster at /private or 1 level below - done with the *
    worker_cluster = "/private/very"
    worker_restrictions = "/private/very || /private/very/+"
    Define a host that will only run jobs in /private/very or any level below - done with the +

    Hosts

    Qube! allows users to specify a list of hosts, for the job to run on. This is a comma-delimited list of hostnames.

    Example
    % qbsub --hosts "qb001,qb002" command


    Omit Hosts

    Qube! allows users to specify a list of hosts on which the job is restricted from running. This is a comma-delimited list of hostnames. This blacklists the hosts from running on these hosts.

    Example
    % qbsub --omithosts "qb001,qb002" command

    Host Groups

    Qube! allows the administrator to organize the farm into clusters or host groups. These groups have no hierarchy, and hosts are allowed membership in multiple groups. In order to restrict a job to a specific set of hosts, the user may specify in the 'group' field of the job, which groups they want the job restricted to.

    Example
    % qbsub --groups "vfx,character" Render my/file.ma

     

    Omit Groups
    This is the opposite of the Host Groups option, where the job will restrict itself from running upon any hosts which are contained by the specified groups.

    Example
    % qbsub --omitgroups "vfx,character" command

    Host Order

    By default Qube! chooses any host (Worker) in the list of hosts which qualify. If given a choice, a job is allowed to prefer a particular host based upon its attributes. This is established using the Qube! resources and priorities defined earlier in the Requirements section of this document.

    Any worker resource or property can be specified, but the most commonly used are:

    • host.processors
    • host.memory
    • host.processor_speed

    Syntax

    [+|-]host.property
    [+|-]host.resource.[total|used|avail]

    The + or  in the expression is used to determine if the job would prefer the largest or smallest value possible.  If neither is used, + is assumed.

    Examples

    CommandMeaning

    % qbsub --hostorder "host.processor_speed" Render myscene.ma

    Choose the fastest host
    % qbsub --hostorder "-host.processors.used" Render myscene.maChoose the host with the least number of worker slots in use
    % qbsub --hostorder "host.processor_speed,host.processors.avail" Render myscene.maChoose the fastest host with the most available worker_slots
    Notes

    The system will use the hostorder specification only when initially choosing the most preferable Worker for the job itself. Once it has chosen a host, it will try to fill it up with instances from the job until the host is full. In other words, the system will not attempt to apply the hostorder to select a host for each individual instance.

    See Also

    supervisor_default_hostorder

    • No labels