Molecular Structure Determination on a Computational and Data Grid - Parallel Computing for Bioinformatics and Computational Biology - page 591

TABLE 24.1 Sample Computational Resource Queue
Snapshot

Job id User Nodes Procs Wall-Time Status

1 user2 32 64 360 Running

2 user1 32 64 360 Running

3 user1 32 64 360 Running

4 user1 32 64 360 Running

5 user1 32 64 360 Running

6 user3 64 128 500 Running

7 user5 34 68 720 Running

8 user4 96 192 720 1

9 user5 64 128 360 2

10 user5 64 128 480 3

11 user5 128 256 720 4

12 user6 128 256 720 5

13 user5 128 256 720 6

14 user6 96 192 306 7

15 user5 64 128 480 8

jobs and eight queued jobs, where the queue job priority determines a relative rank for
corresponding to the order that the queued jobs will start. Note that the user requests
the number of nodes, number of processes, and wall-time queue parameters for each
of the running and queued jobs. This information is sufficient to completely define
the job execution and the native scheduler priority determination.
The native queue scheduler uses the designated job wall-time for managing
resource backfill and estimated job start and end times. Table 24.2 reports the native

TABLE 24.2 Native Queue Job Execution Start and
End Time

Job id Wall-Time Start Time End Time

1 360 00:05:41 00:11:41

2 360 00:05:41 00:11:41

3 360 00:05:41 00:11:41

4 360 00:05:41 00:11:41

5 360 00:05:41 00:11:41

6 500 00:05:41 00:14:01

7 720 00:07:36 00:19:36

8 720 00:11:41 00:23:41

9 360 00:11:41 00:17:41

10 480 00:14:01 00:22:01

11 720 00:23:41 01:11:41

12 720 00:22:01 01:10:01

13 720 01:11:41 01:23:41

14 306 01:10:01 01:15:07

15 480 01:23:41 02:07:41