bsla

service_class_name

The name of a service class that is configured in the lsb.serviceclasses file.

-h

Prints command usage to stderr and exits.

-V

Prints LSF release version to stderr and exits.

-N

Displays information about service class job counts by job slots instead of number of jobs. NSLOTS, and jobs in PEND, RUN, SSUSP, or USUSP state are all counted in slots rather than number of jobs.

Time-based SLAs typically have throughput, velocity, or deadline goals. A list of service classes is displayed with the following fields:

SERVICE CLASS NAME

The name of the service class, followed by its description, if any.

PRIORITY

The service class priority. A higher value indicates a higher priority, relative to other service classes. Similar to queue priority, service classes access the cluster resources in priority order.

USER GROUP

User names or user groups who can submit jobs to the service class.

GOAL

The type of service class goal and its configured value.

ACTIVE WINDOW

The configured time window when the service class goal is active. If no time window is configured for a throughput or velocity goal, the ACTIVE WINDOW is Always Open.

STATUS

Status of the service class goal.

• Active:On time means that the goal is active and meeting its target.
• Active:Delayed means that the goal is active but is missing its target.
• Inactive means that the goal is not active, and that its time window is closed. Jobs are scheduled as if no service class is defined. LSF does not enforce any service-level goal for an inactive SLA.

THROUGHPUT

For throughput goals, the configured job throughput (finished jobs per hour) for the service class.

SLA THROUGHPUT

The current throughput for the SLA finished jobs per clean period.

ESTIMATED FINISH TIME

For goals with a time window, estimated finish time of the SLA. If the service class status is on time, the finish time is before the configured deadline. If the service class status is delayed, the service class is missing its goal and the bsla command shows a finish time later than the deadline.

OPTIMUM NUMBER OF RUNNING JOBS

For goals with a time window, the optimum number of jobs that must be running in the service class for the SLA to meet its goal.

NJOBS

The current number of jobs in the specified service class. A parallel job is counted as one job, regardless of the number of job slots it uses.

PEND

The number of pending jobs in the specified service class.

RUN

The number of running jobs in the specified service class.

SSUSP

The number of system-suspended jobs in the service class.

USUSP

The number of user-suspended jobs in the specified service class.

FINISH

The number of jobs in the specified service class in EXIT or DONE state.

SERVICE CLASS NAME

The name of the service class, followed by its description, if any.

GOAL

The type of service class goal and its configured value (GUARANTEE).

AUTO_ATTACH

Automatic attachment configuration (Y or N).

ACCESS_CONTROL

Configured access restrictions for the guarantee SLA, if any.

POOL NAME

Name of the guaranteed resource pool.

TYPE

Guaranteed resource type.

GUAR CONFIG

Number of resources in the pool that is guaranteed to the SLA.

GUAR USED

Number of resources within the guarantee in use by the SLA. Resource use includes both running and suspended jobs.

TOTAL USED

Number of resources in the pool currently in use by the SLA. This total can exceed the number of guaranteed resources for the SLA if other guaranteed SLAs that use the same resource pool are not running at capacity. Resource use includes both running and suspended jobs.

USED GUARANTEE HOSTS

Information on the hosts that are allocated from each guarantee pool for the SLA, organized by guarantee pool. This section only displays if there are jobs running in the SLA.

CONSUMER

The name of the EGO consumer from which hosts are allocated to the SLA.

EGO_RES_REQ

The EGO resource requirement that is defined in the SLA.

MAX_HOST_IDLE_TIME

How long the SLA holds its idle hosts before LSF releases them to EGO.

NUM_RECALLED_HOSTS

The number of hosts that are allocated to the SLA that EGO reclaimed.

RECALLED_HOSTS_TIMEOUT

The amount of time EGO gives to LSF to clean up its workload before EGO reclaims the host.

The following time-based service class that is named Duncan is configured in the lsb.serviceclasses file:

Begin ServiceClass
NAME = Duncan
CONSUMER = Duncan
PRIORITY = 23
USER_GROUP = user1 user2
GOALS = [VELOCITY 8 timeWindow (9:00-17:30)] \ 
[DEADLINE timeWindow (17:30-9:00)] 
DESCRIPTION = Daytime/Nighttime SLA
End ServiceClass

The bsla command shows the following properties and status:

bsla Duncan 
SERVICE CLASS NAME:  Duncan
 -- Daytime/Nighttime SLA
PRIORITY:  23
CONSUMER: Duncan
EGO_RES_REQ: any host
MAX_HOST_IDLE_TIME: 120
USER_GROUP:  user1 user2
 
GOAL:  VELOCITY 8
ACTIVE WINDOW: (9:00-17:30) 
STATUS:  Active:On time
SLA THROUGHPUT:  0.00 JOBS/CLEAN_PERIOD
GOAL:  DEADLINE 
ACTIVE WINDOW: (17:30-9:00) 
STATUS:  Inactive
SLA THROUGHPUT:  0.00 JOBS/CLEAN_PERIOD
 
 NJOBS   PEND    RUN     SSUSP   USUSP   FINISH
      0      0       0        0       0       0

The following resource pools that are named linuxPool and solarisPool are configured in the lsb.resources file:

Begin GuaranteedResourcePool
NAME =linuxPool
TYPE = hosts
HOSTS = linuxHG
DISTRIBUTION = [[sla1,10%] [sla2,25%]
DESCRIPTION = A linux resource pool used by sla1, and sla2.
End GuaranteedResourcePool
 

Begin GuaranteedResourcePool
NAME =solarisPool
TYPE = hosts
HOSTS = solarisHG
DISTRIBUTION = [[sla1,20%] [sla2,30%] [sla3,25%]]
DESCRIPTION = A solaris resource pool used by sla1, sla2, and sla3.
End GuaranteedResourcePool

The bsla command shows the following information for service class sla1:

bsla sla1
SERVICE CLASS NAME:  sla1
 -- SLA ONE
ACCESS CONTROL:    QUEUES[normal]  FAIRSHARE__GROUPS[lsfadmins/]
AUTO ATTACH:       Y
GOAL:              GUARANTEE                    GUARANTEE    GUARANTEE    TOTAL
POOL NAME      TYPE    CONFIG         USED     USEDslotPool       slots       10            0        0