User Tools

Site Tools


cluster:167

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
Next revision Both sides next revision
cluster:167 [2018/06/28 14:02]
hmeij07
cluster:167 [2018/06/28 14:09]
hmeij07
Line 25: Line 25:
 | Job Hours | 221,136  |  77:1  |  2,872 | cumulative hours of consumed usage | | Job Hours | 221,136  |  77:1  |  2,872 | cumulative hours of consumed usage |
 | Job Hours % | 31  |  6:1  |  5 | as a percentage of available | | Job Hours % | 31  |  6:1  |  5 | as a percentage of available |
-| Avail Hours2 | 561,600  |  39:1  |  14,400 | total cpu cores hp12's 256 cores, total gpus | +| Avail Hours2 | 561,600  |  39:1  |  14,400 | total cpu cores minus hp12's 256 cores, total gpus | 
-| Job Hours2 % | 39  |  8:1  |  5 | more realistic...hp12 rarely used in June18|+| Job Hours2 % | 39  |  8:1  |  5 | more realistic...hp12 rarely used |
  
 The logs showing gpu %util confirm the extremely low GPU usage. When concatenating the four gpu %util values into a string, since 01Jan2017, the string '0000' has occurred 10 million times out of 16 million observations. (GPUs are polled every 10 mins). Sad. The surprising strong GPU job count is due to the Amber group launching lots of small GPU jobs. The logs showing gpu %util confirm the extremely low GPU usage. When concatenating the four gpu %util values into a string, since 01Jan2017, the string '0000' has occurred 10 million times out of 16 million observations. (GPUs are polled every 10 mins). Sad. The surprising strong GPU job count is due to the Amber group launching lots of small GPU jobs.
  
-So were these 25 days in June 2018 an oddity? March is Honors' Theses time so lets look at Jul17 so we can compare that to Jul18 in august.+So were these 25 days in June 2018 an oddity? 
  
 ^  Total Monthly CPU+GPU Hours  ^^^^^^^^^^^ ^  Total Monthly CPU+GPU Hours  ^^^^^^^^^^^
 ^Ju17^Aug17^Sep17^Oct17^Nov17^Dec17^Jan18^Feb18^Mar18^Apr18^May18^ ^Ju17^Aug17^Sep17^Oct17^Nov17^Dec17^Jan18^Feb18^Mar18^Apr18^May18^
 |313,303|273,051|128,390|111,224|280,101|51,727|306,453|222,585|437,959|262,227|294,724| |313,303|273,051|128,390|111,224|280,101|51,727|306,453|222,585|437,959|262,227|294,724|
 +
 +March is Honors' Theses time so lets look at Jul17 so we can compare that to Jul18 in august. 31 days in July is 744 hours.
  
 ^ Metric ^ CPU ^ Ratio ^ GPU ^ Notes July 2017 ^ ^ Metric ^ CPU ^ Ratio ^ GPU ^ Notes July 2017 ^
Line 46: Line 48:
 | Job Hours % | 30  |  1:1  |  26 | as a percentage of available | | Job Hours % | 30  |  1:1  |  26 | as a percentage of available |
 | Avail Hours2 | 696,384  |  47:1  |  14,880 | total for cpu cores minus hp12's 256 cores, total for gpus | | Avail Hours2 | 696,384  |  47:1  |  14,880 | total for cpu cores minus hp12's 256 cores, total for gpus |
-| Job Hours2 % | 37  |  1.5:1  |  26 | more realistic...hp12 rarely used in June18|+| Job Hours2 % | 37  |  1.5:1  |  26 | more realistic...hp12 rarely used |
  
   * Some noise in this data with the inability to match start and end of job (~15% of records)   * Some noise in this data with the inability to match start and end of job (~15% of records)
cluster/167.txt · Last modified: 2018/08/01 14:11 by hmeij07