| COLUMN | DATA TYPE | DESCRIPTION | EXAMPLE |
|---|
| JOB_NAME | VARCHAR(64) | | 207777.mira |
| COBALT_JOBID | INTEGER | COBALT Job ID | 207777 |
| MACHINE_NAME | VARCHAR(64) | Machine Name | mira |
| AWARD_CATEGORY_NAME | VARCHAR(64) | The award program for the allocation for the job. INCITE, ALCC, DISCRETIONARY | mira |
| QUEUED_TIMESTAMP | TIMESTAMP | Timestamp when a job entered the queue for a machine | 2014-01-28 01:22:44.029342 |
| QUEUED_DATE_ID | INTEGER | DATE_ID of when job entered the queue for a machine | 20140128 |
| START_TIMESTAMP | TIMESTAMP | Timestamp when a job started | 2014-01-28 13:22:44.029342 |
| START_DATE_ID | INTEGER | DATE_ID of when a job started | 20140128 |
| END_TIMESTAMP | TIMESTAMP | Timestamp when a job ended | 2014-01-28 17:22:44.029342 |
| END_DATE_ID | INTEGER | DATE_ID of when a job ended | 20140128 |
| USERNAME_GENID | VARCHAR(64) | An integer value in place of the user name. Other records with the same value means they have the same user name. | 123456789012 |
| PROJECT_NAME_GENID | VARCHAR(64) | An integer value in place of the project name. Other records with the same value means they have the same project name | 987654321012 |
| QUEUE_NAME | VARCHAR(64) | Queue Name | prod-short |
| SCIENCE_FIELD | VARCHAR(64) | Science Field: Sub Science field. The main science category with the sub category. Internal,Support means it is an internal or support project used by ALCF staff. | Physics: Nuclear Physics |
| SCIENCE_FIELD_SHORT | VARCHAR(64) | Science Field. The main science category without the sub category. Internal,Support means it is an internal or support project used by ALCF staff. | Physics |
| WALLTIME_SECONDS | DECIMAL(24,4) | The time requested for a job, can be different than runtime | 14400 |
| RUNTIME_SECONDS | DECIMAL(24,4) | The elapsed time the job executed. | 12590 |
| NODES_USED | DECIMAL(24,2) | Number of nodes used as determined by the scheduler. Can be different than NODES_REQUESTED. If the NODES_REQUESTED is 500, the job could be assigned 512 nodes. | 512 |
| NODES_REQUESTED | DECIMAL(24,2) | Number of nodes requested, can be different than the NODES_USED value. | 499 |
| CORES_USED | DECIMAL(24,2) | Number of cores used, NODES_USED * CORES PER NODE | 8192 |
| CORES_REQUESTED | DECIMAL(24,2) | Number of cores used, CORES_REQUESTED * CORES PER NODE | 7984 |
| LOCATION | VARCHAR(2048) | Location on machine, syntax will vary depending on machine | MIR-480C0-7B3F1-512 |
| EXIT_STATUS | INTEGER | Unix process exit status | 0 |
| ELIGIBLE_WAIT_SECONDS | INTEGER | The time a job waited in the queue, minus any time where the job was not eligible to run, such as being in a hold. | 41256 |
| ELIGIBLE_WAIT_FACTOR | INTEGER | The ELIGIBLE_WAIT_SECONDS over the wallclock time requested. | 3 |
| QUEUED_WAIT_SECONDS | INTEGER | The time a job waited in queue. This includes time held by the user and other ineligible times. | 43200 |
| QUEUED_WAIT_FACTOR | INTEGER | The QUEUED_WAIT_SECONDS over the wallclock time requested. | 3 |
| REQUESTED_CORE_HOURS | DECIMAL(24,4) | Requested core hours | 31936 |
| USED_CORE_HOURS | DECIMAL(24,4) | Used core hours | 32768 |
| CAPABILITY_USAGE_CORE_HOURS | DECIMAL(26,4) | If job is Capability, the hours are here | 0 |
| NONCAPABILITY_USAGE_CORE_HOURS | DECIMAL(26,4) | If job is Noncapability, the hours are here | 32768 |
| BUCKETS3_A_USAGE_CORE_HOURS | DECIMAL(26,4) | If job is Bucket A the hours are here | 0 |
| BUCKETS3_B_USAGE_CORE_HOURS | DECIMAL(26,4) | If job is Bucket B the hours are here | 0 |
| BUCKETS3_C_USAGE_CORE_HOURS | DECIMAL(26,4) | If job is Bucket C the hours are here | 32768 |
| MACHINE_PARTITION | VARCHAR(4064) | The value for a mapping of midplanes on the machine. | MIR-08840-3BB71-512 |
| EXIT_CODE | INTEGER | UNIX exit code of the executable for the job. | 0 |
| MODE | VARCHAR(64) | The qsub command mode flag. Valid modes on the EAS are c1, c2, c4, c8, c16, c32, c64, script and interactive. Script is a script invocation as it was on the BG/P systems. The cN modes correspond to -p N being used with runjob where N is the
number of ranks. Interactive is for running an job from the console. | c1 |
| RESID | INTEGER | Cobalt reservation id, if the job was run within a reservation. | |
| DATA_LOAD_STATUS | VARCHAR(16) | complete,incomplete | complete |
| CAPABILITY | VARCHAR(16) | Calculated Capabiltiy | Not capablity |
| SIZE_BUCKETS3 | VARCHAR(64) | Calculated Job Size Bucket | 0% <= x < 16.7% |
| PERCENTILE | VARCHAR(64) | Calculated Percentile Size | 0% <= x < 16.7% |
| NUM_TASKS_SUBBLOCK | INTEGER | Number of subblock tasks: if a job has multiple tasks and multiple tasks fit in a midplane and they run simultaneously. | 0 |
| NUM_TASKS_CONSECUTIVE | INTEGER | Number of consecutive tasks: if a job has multiple tasks and they are run one after another, it is consecutive. | 0 |
| NUM_TASKS_MULTILOCATION | INTEGER | Number of multilocation tasks. If a job runs multiple tasks simultaneously, it is a multilocation job. | 0 |
| NUM_TASKS_SINGLE | INTEGER | Number of single tasks: If a job runs one task, it is a single task job. | 1 |
| COBALT_NUM_TASKS | INTEGER | Number of tasks recorded by cobalt | 1 |
| IS_SINGLE | SMALLINT | Is Single Task Job | 1 |
| IS_CONSECUTIVE | SMALLINT | if a job has multiple tasks and they are run one after another, 1, else 0 | 0 |
| IS_MULTILOCATION | SMALLINT | If a job runs multiple tasks simultaneously, 1 else 0 | 0 |
| IS_SUBBLOCK | SMALLINT | If a job has multiple tasks and multiple tasks fit in a midplane and they run simultaneously, 1 else 0 | 0 |
| IS_SUBBLOCK_ONLY | SMALLINT | If a job has multiple tasks and multiple tasks fit in a midplane and they run simultaneously and tasks are not consecutive, 1 else 0 | 0 |
| IS_MULTILOCATION_ONLY | SMALLINT | If a job runs multiple tasks simultaneously and not subblock, 1 else 0 | 0 |
| IS_MULTILOCATION_SUBBLOCK | SMALLINT | If a job runs multiple tasks simultaneously and is subblock, 1 else 0 | 0 |
| IS_CONSECUTIVE_ONLY | SMALLINT | if a job has multiple tasks and they are run one after another and not subblock or multilocation, 1, else 0 | 0 |
| IS_SINGLE_ONLY | SMALLINT | If is a Single Task Job and not subblock or consecutive, 1 else 0 | 0 |
| IS_NO_TASKS | SMALLINT | If job has no tasks, 1 else 0 | 0 |
| IS_OTHER | SMALLINT | If job is not in any other category | 0 |
| IS_DAOS | SMALLINT | If job used DAOS system | 0 |
| OVERBURN_CORE_HOURS | DECIMAL(24,4) | If job is capability and the total hours of all jobs for the allocation up to the timestamp this job ran is greater than the allocation amount then the hours are overburn hours. | 0 |
| IS_OVERBURN | SMALLINT | If the job has overburn hours, then 1. A job is overburn if it is capability and the total hours of all jobs to the time this job ran is greater than the allocation amount. | 0 |
| GPUS_REQUESTED | INTEGER | GPUs Requested | |
| GPUS_USED | INTEGER | GPUs Used based on GPU utilization of at least 10% for each GPU on all of the nodes used during job start and end times | |
| GPU_UTILIZATION | DECIMAL(12,8) | Average utilization value from 0.0 to 100.0 of all job's GPUs used during the job run, GPUs whose load was less than 10% are not including in this average value. For nvidia-smi it is utilization.gpu: Percent of time over the past sample period during which one or more kernels was executing on the GPU. | |
| GPU_MEMORY_ALLOCATED_AVERAGE | BIGINT | Average of the max memory allocated per GPU of all job's GPUs used during the job run, GPUs whose load was less than 10% are not including in this average value. For nvidia-smi it is the memory.used: Total memory allocated by active contexts. Units: KiB (kibibytes) | |
| GPU_MEMORY_UTILIZATION_AVERAGE | DECIMAL(12,8) | Average memory utilization value from 0.0 to 100.0 of all job's GPUs used during the job run, GPUs that were whose load was less than 10% are not including in this average value. For nvidia-smi it is utilization.memory: Memory utilization refers to the percentage of time over the past sampling period during which the GPU's memory subsystem was actively being read from or written to. | |
| GPU_POWER_USAGE_AVERAGE | DECIMAL(12,8) | Average power usage in watts of all GPUs for the nodes for the job. It is the average sampled value of all GPUs of the nodes used by the job for the time period. For nvdia-smi it is power.draw: measured power draw for the entire board, in watts | |
| GPU_TEMPERATURE_AVERAGE | DECIMAL(12,8) | Average temperature of all GPUs for the nodes for the job. It is the average sampled value of all GPUs of the nodes used by the job for the time period. For nvdia-smi it is temperature.gpu: Core GPU temperature. in degrees C | |
| GPUS_USED_PER_NODE_AVERAGE | DECIMAL(6,4) | A value from 0.0 to Num GPUs per node. It is the average number of GPUs per node that were utilized. For example a value of 2.0 for a job on the polaris machine, with 4 GPUs per node, would mean half of the requested GPUs had an average utilization of at least 10%. | |
