...
- inventory of existing hardware
- model for retirement vs time
- model for project needs vs time
A python script has been developed to do the modeling. We are using "CPU factor" as the computing unit to account for differing oomphs of the various node types in the farm.
Purchase Record of Existing Hardware
...
In practice, we may adopt a hybrid of these two, especially since a strict age cutoff would make sudden drops in capacity, given our acquisition history.
There are industry estimates for survival rates vs time.
Age cut:
Using a strict 5 year cut, here is the survival rate of the existing hardware (in 2019 it is all gone):
Year | #hosts | #cores | kiso-units | SLAC-units |
---|---|---|---|---|
2013 | 729 | 7164 | 7433 | 90698 |
2014 | 427 | 4848 | 5366 | 65476 |
2015 | 179 | 2864 | 3284 | 40067 |
2016 | 179 | 2864 | 3284 | 40067 |
2017 | 179 | 2864 | 3284 | 40067 |
2018 | 179 | 2864 | 3284 | 40067 |
Basically by the end of 2014, all hardware before the bullet purchase would have been retired.
DNR model:
None are allowed after 10 years.
Year | #hosts | #cores | kiso-units | SLAC-units |
---|---|---|---|---|
2013 | 729 | 7164 | 7433 | 90698 |
2014 | 652 | 6568 | 6862 | 83744 |
2015 | 575 | 5968 | 6288 | 76736 |
2016 | 499 | 5376 | 5718 | 69789 |
2017 | 420 | 4792 | 5166 | 63039 |
2018 | 337 | 3944 | 4275 | 52183 |
2019 | 203 | 2688 | 3024 | 36908 |
2020 | 110 | 1760 | 2018 | 24622 |
2021 | 87 | 1392 | 1596 | 19474 |
2022 | 67 | 1072 | 1229 | 14997 |
2023 | 49 | 784 | 899 | 10968 |
Needs Estimation
PPA projects were polled for their projected needs over the next few years. This is recapped here.