Cockroach Labs Estate economics What idle nodes, provisioned bytes and peak-sized bills cost, worked out from your estate.

Estate summary

  1. Nodes 150 baseline
  2. Avg CPU utilization 15.0% idle hardware
  3. Storage provisioned 3.0 TB 500 GB of data
  4. Est. monthly saving 0%

Step 1 of 8 - Virtualization

You're running an estate, not a cluster

Every cluster below was sized for its own peak, so every one of them idles.

50 standalone clusters, 150 nodes in totalcluster-0110% CPU3 nodescluster-025% CPU3 nodescluster-0321% CPU3 nodescluster-0417% CPU3 nodescluster-059% CPU3 nodescluster-0614% CPU3 nodescluster-0710% CPU3 nodescluster-0817% CPU3 nodescluster-0914% CPU3 nodescluster-1012% CPU3 nodescluster-1110% CPU3 nodescluster-1218% CPU3 nodescluster-1319% CPU3 nodescluster-1424% CPU3 nodescluster-1512% CPU3 nodescluster-1613% CPU3 nodescluster-1717% CPU3 nodescluster-1824% CPU3 nodescluster-1913% CPU3 nodescluster-2014% CPU3 nodescluster-2112% CPU3 nodescluster-2216% CPU3 nodescluster-2319% CPU3 nodescluster-2415% CPU3 nodescluster-2518% CPU3 nodescluster-2618% CPU3 nodescluster-2711% CPU3 nodescluster-2828% CPU3 nodescluster-2912% CPU3 nodescluster-3023% CPU3 nodescluster-3121% CPU3 nodescluster-321% CPU3 nodescluster-3315% CPU3 nodescluster-349% CPU3 nodescluster-3518% CPU3 nodescluster-3614% CPU3 nodescluster-3715% CPU3 nodescluster-3814% CPU3 nodescluster-3915% CPU3 nodescluster-4028% CPU3 nodescluster-418% CPU3 nodescluster-4214% CPU3 nodescluster-4322% CPU3 nodescluster-4414% CPU3 nodescluster-4513% CPU3 nodescluster-4610% CPU3 nodescluster-4713% CPU3 nodescluster-4820% CPU3 nodescluster-4911% CPU3 nodescluster-5010% CPU3 nodesEstate before consolidation50 separate clusters of around 3 nodes each, 150 nodes in total, averaging 15.0% CPU utilization.
15%Average CPUAverage CPU CPU utilization15.0% of provisioned CPU is in use.

150 nodes running 180 vCPU of actual work.

The arithmetic, in full
StepWorkingResult
Provisioned today50 clusters of 3 nodes each, 150 in total x 8 vCPU1200 vCPU
Mean demand1200 vCPU x 15.0% estate average utilization (per-cluster 0.5% to 28.0%, sd 5.3%)180 vCPU
Aggregate peak180.0 vCPU x (1 + (2.23 - 1) / sqrt(50)), a pooled peak-to-mean of 1.174211 vCPU
Consolidated, sized once for the peakmax(3, ceil(211.3 vCPU / (8 vCPU x 37.5% target))) = max(3, ceil(70.42)); the mean sits well under that peak71 nodes, averaging 31.7%
Auto-scaled through the daythe pool follows the pooled curve in whole nodes, 30 minutes behind it, holding its size for 2 hours before shrinking62 nodes on average, at 36.3% CPU
Utilization across the day31.9% to 38.7% against a 37.5% setpoint: the pool runs hot while it's a bucket behind a rise, and cool while the cooldown is still holding capacity after a peak31.9%-38.7%

Assumptions

  • 8 vCPU per node. Standalone clusters are drawn around 3 nodes, snapped to an odd number for quorum and held between 3 and 15; in this estate they're all 3, because a spread centred on the end of that range has nowhere to go, for 150 nodes in total.
  • Each standalone cluster's CPU is a draw from a normal distribution centred on 15.0% and held between 0% and 100%; in this estate they run 0.5% to 28.0%, a 5.3-point standard deviation. The published band is 10-15% today against a 30-45% auto-scaled target, and this Private Host Cluster is set to 37.5%.
  • The Private Host Cluster floors at 3 nodes and must reach 71 at peak.
  • The auto-scaler sizes for what it saw 30 minutes ago and waits 2 hours before shrinking, so the pool averages 36.3% rather than sitting on its 37.5% setpoint.
  • Workloads are independent; consolidation moves where work runs, not how much there is.
What a virtual cluster gives an app team

Its own connection string, databases, schemas, users, roles and backups, and no visibility into the Private Host Cluster or any other virtual cluster.