Search
Page 28 of 40. Showing 393 results (0.013 seconds)
execute.py
= qb.jobobj() if bool(subjob['package'].get('dev')): pp = pprint.PrettyPrinter(indent=2, width=1) print "JOB:" pp.pprint(subjob) return subjob def…7. Final Page
. On Linux, this will be the root user. Installed a worker: You will have 3 test buttons: test simple job - this will test basic operation of this worker on your farm. Job status…User_Proxy mode Considerations
By default, each Qube Worker is configured to run in what we call "proxy" mode. This means the Worker will attempt to execute the job as a proxy user, rather than the user…Worker Host Groups
Host groups are a simple way to target jobs to a list of machines without specifying them explicitly every time you submit a job. To set up a host group, the Worker needs…Queuing Algorithm
Despite the number of processors in a render farm, it is still necessary to order the jobs so that they are dispatched to execution hosts in a deterministic order. Queuing…supervisor_manifest
Synopsis supervisor_manifest = path path were the manifest log is located. This log contains all events concerning jobs and their actions. The format is similar to the job…supervisor_preempt_policy
preemption routine will determine if the system should kill a job and replace it. However, the administrator is given the flexibility to determine if they want their jobs to die…supervisor_max_worker_requeue
Synopsis supervisor_max_worker_requeue = retries Retries is the maximum number of attempts the Supervisor will make to assign a job to a Worker host that claims to have…supervisor_verbosity
Synopsis supervisor_verbosity = int or string Flags mask to determine what information to go into the supervisor log. Values startupwork subjob job host log admin…proxy_account
Synopsis proxy_account = user This is only to be specified if the administrator wants all jobs to be run under the single user account. By default, all jobs are run…