Cockroach Labs Estate economics What idle nodes, provisioned bytes and peak-sized bills cost, worked out from your estate.
  1. Nodes 300 baseline
  2. Avg CPU utilization 15.0% idle hardware
  3. Storage provisioned 6.0 TB 1.0 TB of data
  4. Est. monthly saving 0%

Step 7 of 8 - Pricing

You pay for the area, not the rectangle

Provisioned capacity is a flat ceiling sized to the busiest fifteen minutes of the day, plus headroom for the peak you didn't forecast. The gap under that ceiling is hardware you rented and didn't use.

Run a day, 40% headroom Steady Business hours Batch spike

Provisioned capacity left unused

not started

Pick a day above. The ceiling is already drawn: it's sized to a peak of 802 vCPU plus 40%, and you pay for all of it whether the workload turns up or not. That's under the 2,400 vCPU the estate runs on today, which was bought a cluster at a time and works at 15.0%, not sized to this curve.

0561112300:0006:0012:0018:0024:00vCPUProvisioned1123 vCPUConsumed360 vCPU avgProvisioned capacity versus consumption over 24 hoursOver 24 hours, a ceiling of 1123 vCPU, set above a peak of 802 vCPU, costs 26950 vCPU-hours, while the workload actually consumes 8640 vCPU-hours: 68% of the provisioned capacity goes unused.
Provisioned
0 vCPU-hours
Consumed
0 vCPU-hours
Paid for, not used
0 vCPU-hours
Read this chart as a table
Mean demand per hour against provisioned capacity, in vCPU
HourDemandProvisionedUnused
00:00 163 1,123 960
01:00 163 1,123 960
02:00 163 1,123 960
03:00 162 1,123 961
04:00 164 1,123 959
05:00 166 1,123 957
06:00 162 1,123 961
07:00 223 1,123 900
08:00 376 1,123 747
09:00 521 1,123 602
10:00 638 1,123 485
11:00 719 1,123 403
12:00 776 1,123 347
13:00 780 1,123 343
14:00 743 1,123 380
15:00 669 1,123 454
16:00 557 1,123 566
17:00 414 1,123 709
18:00 265 1,123 858
19:00 162 1,123 961
20:00 165 1,123 958
21:00 163 1,123 960
22:00 163 1,123 960
23:00 164 1,123 959
How this is worked out
  • Demand is this estate's 360 vCPU mean, shaped by the preset you picked, which also drives step 1.
  • Capacity is sized at the day's peak of 802 vCPU plus the 40% headroom you picked, giving a ceiling of 1,123 vCPU. The headroom is an assumption rather than a published figure, so it's yours to argue with.
  • That peak is the 100 clusters' own peaks added up, because a provisioned estate buys each of them separately. Step 1's aggregate peak of 404 vCPU is the later, smaller question: what one pool needs once the workloads share it.
  • A ceiling that high averages 32.1% CPU across the day, which is the whole day's 67.9% of unused capacity counted the other way round. The pooled cluster on step 1 runs at 36.9% against a 37.5% setpoint. Neither number sets the other: this one is how far above the peak you buy, that one is what a controller is told to hold.
Where the storage number comes from

3 Raft copies at 50% utilization (3/0.5 = 6), on top of the block store's own internal replication.

1.0 TB of logical data becomes 6.0 TB provisioned , a factor of 6× before the cost model sees it.