PDSF Home
Getting StartedAccount Request
Access
Startup Files
Passwords
Cluster Stats
Ganglia
Batch Stats
Network Stats
Internal
External
Batch Summary
Storage Stats
Help
Bug Tracking
FAQ
Support Request
Group Pages
Star
Atlas
ICECUBE
Hardware
Ganglia Moved to Monitoring
File System
Disk Vaults
Compute Nodes
Mass Storage
Networking
Announcements
Email Lists
Meetings
Schedule
News & Status
Message of the DayNERSC
Announcements
PDSF LSF Batch Performance from January 1, 2002 to
less than February 1, 2002
Number of CPU's available during the accounting period:
Number of CPU days available on the cluster: 31 (days) * 0 (nodes) = 0 days
Number of jobs run during that time: 1278
Wall clock time used by the jobs: 78.69 days (0.00% utilization)
CPU time used by the jobs: 66.10 days (83.99% efficiency)
| Date | Wall | Used | Faults | Count | ||||
|---|---|---|---|---|---|---|---|---|
| Seconds | Hours | Total % | Seconds | Hours | % Efficiency | |||
| 1 | 4084841.00 | 1134.68 | 60.08% | 3700411.92 | 1027.89 | 90.59% | 208.17 | 803 |
| 2 | 746034.00 | 207.23 | 10.97% | 688702.16 | 191.31 | 92.32% | 128.01 | 114 |
| 3 | 15134.00 | 4.20 | 0.22% | 13942.14 | 3.87 | 92.12% | 181.20 | 0 |
| 4 | 1371188.00 | 380.89 | 20.17% | 1269726.97 | 352.70 | 92.60% | 146.52 | 323 |
| 19 | 59830.00 | 16.62 | 0.88% | 1075.59 | 0.30 | 1.80% | 4473.79 | 9 |
| 20 | 91199.00 | 25.33 | 1.34% | 453.53 | 0.13 | 0.50% | 3960.52 | 3 |
| 21 | 17494.00 | 4.86 | 0.26% | 967.31 | 0.27 | 5.53% | 3250.84 | 2 |
| 24 | 91774.00 | 25.49 | 1.35% | 12529.58 | 3.48 | 13.65% | 166.05 | 6 |
| 22 | 17723.00 | 4.92 | 0.26% | 21.30 | 0.01 | 0.12% | 4771.82 | 0 |
| 23 | 86399.00 | 24.00 | 1.27% | 103.86 | 0.03 | 0.12% | 4771.82 | 0 |
| 25 | 19001.00 | 5.28 | 0.28% | 17673.81 | 4.91 | 93.02% | 105.33 | 2 |
| 29 | 35888.00 | 9.97 | 0.53% | 1233.56 | 0.34 | 3.44% | 689.48 | 9 |
| 30 | 76080.00 | 21.13 | 1.12% | 2423.43 | 0.67 | 3.19% | 1296.99 | 6 |
| 31 | 86399.00 | 24.00 | 1.27% | 1378.60 | 0.38 | 1.60% | 779.91 | 1 |
The Efficiency number may be greater than 100% if the processes used threads. SGE aggregated all thread cpu time together, thus efficiency is based on cpu time verses wall clock time.
