Platform Analytics 7 Dataflow

DPR_BYINTERVAL table

This table is being used to store the sampled LSF Jobs of Pending Reason data . The default sampling time is 15 minutes.
Column Name Data Type Nullable (Y/N) Key (PK/FK) Description
CLUSTER_NAME VARCHAR2(128) N PK The cluster name.
TIME_STAMP TIMESTAMP N   The time of this record in the local server time zone.
TIME_STAMP_GMT NUMBER(13) N PK The time of this record in GMT time. This is the number of seconds from midnight, 1 January, 1970.
HOST_TYPE VARCHAR2(128) N PK The type of the host.
USER _NAME VARCHAR2(128) N PK The user who submitted the job.
QUEUE_NAME VARCHAR2(128) N PK The name of the job queue to which the job belongs.
PROJECT_NAME VARCHAR2(128) N PK The project the job was submitted from.
PENDING_REASON_CODE NUMBER(10) N PK The reason the job is in the PEND or PSUSP state.
PENDING_REASON_TYPE VARCHAR2( 32 ) N PK

�Job Related Reasons�;

"Queue and System Related Reasons";

"User Related Reasons";

"Host Related Reasons" ;

�MC Related Reasons�

�Other Reasons�
JOB_NUMBER NUMBER Y   The job number is obtained by counting the number of distinct job IDs grouped by the same ClusterName, PendReasonType, PendReasonCode, UserName, ProjectName, QueueName field values across the same interval.
PEND_MINUTES NUMBER Y  

The total time the jobs was pending in minutes. If the pend hours is less than 0, the minutes is 0. And if the pend hours is less than sampling interval the minutes is the value. If the pend hours is more than sampling interval , the minutes is the interval. Finally, sum the whole minutes of each job to pend minustes

e.g

Jobid reason minutes

1 2 4

2 2 5

The PEND_MINUTES is 4+5=9
INSERT_SEQ NUMBER(15) Y   The order of sequence in which this record is inserted into the table. This field is auto-incremented whenever a new record is inserted into the table.