Cockroach Labs Estate economics What idle nodes, provisioned bytes and peak-sized bills cost, worked out from your estate.
  1. Nodes 650 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.

Deployment model
Separate clusters One Private Host Cluster
50 standalone clusters, 650 nodes in totalcluster-0110% CPU13 nodescluster-025% CPU13 nodescluster-0321% CPU13 nodescluster-0417% CPU13 nodescluster-059% CPU13 nodescluster-0614% CPU15 nodescluster-0710% CPU11 nodescluster-0817% CPU13 nodescluster-0914% CPU13 nodescluster-1012% CPU13 nodescluster-1110% CPU13 nodescluster-1218% CPU13 nodescluster-1319% CPU13 nodescluster-1424% CPU13 nodescluster-1512% CPU13 nodescluster-1613% CPU13 nodescluster-1717% CPU13 nodescluster-1824% CPU13 nodescluster-1913% CPU13 nodescluster-2014% CPU13 nodescluster-2112% CPU15 nodescluster-2216% CPU13 nodescluster-2319% CPU13 nodescluster-2415% CPU15 nodescluster-2518% CPU13 nodescluster-2618% CPU13 nodescluster-2711% CPU13 nodescluster-2828% CPU13 nodescluster-2912% CPU13 nodescluster-3023% CPU13 nodescluster-3121% CPU11 nodescluster-321% CPU13 nodescluster-3315% CPU13 nodescluster-349% CPU13 nodescluster-3518% CPU13 nodescluster-3614% CPU13 nodescluster-3715% CPU13 nodescluster-3814% CPU13 nodescluster-3915% CPU11 nodescluster-4028% CPU13 nodescluster-418% CPU13 nodescluster-4214% CPU13 nodescluster-4322% CPU13 nodescluster-4414% CPU13 nodescluster-4513% CPU13 nodescluster-4610% CPU13 nodescluster-4713% CPU11 nodescluster-4820% CPU13 nodescluster-4911% CPU13 nodescluster-5010% CPU15 nodesEstate before consolidation50 separate clusters of around 13 nodes each, 650 nodes in total, averaging 15.0% CPU utilization.
15%Average CPUAverage CPU CPU utilization15.0% of provisioned CPU is in use.

650 nodes running 780 vCPU of actual work.

The arithmetic, in full
StepWorkingResult
Provisioned today50 clusters totalling 650 nodes (drawn around 13, 11 to 15) x 8 vCPU5200 vCPU
Mean demand5200 vCPU x 15.0% estate average utilization (per-cluster 0.5% to 28.1%, sd 5.3%)780 vCPU
Aggregate peak780.0 vCPU x (1 + (2.23 - 1) / sqrt(50)), a pooled peak-to-mean of 1.174915 vCPU
Consolidated, sized once for the peakmax(3, ceil(915.5 vCPU / (8 vCPU x 37.5% target))) = max(3, ceil(305.15)); the mean sits well under that peak306 nodes, averaging 31.9%
Auto-scaled through the daythe pool follows the pooled curve in whole nodes, 30 minutes behind it, holding its size for 2 hours before shrinking266 nodes on average, at 36.7% CPU
Utilization across the day32.2% to 38.9% 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 peak32.2%-38.9%

Assumptions

  • 8 vCPU per node. Standalone clusters are drawn around 13 nodes, snapped to an odd number for quorum and held between 3 and 15; in this estate they run 11 to 15, a 0.8-node standard deviation, for 650 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 28.1%, 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 306 at peak.
  • The auto-scaler sizes for what it saw 30 minutes ago and waits 2 hours before shrinking, so the pool averages 36.7% 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.