# Shared VPS plans with the worst CPU steal spikes

Shared-CPU plans ranked by their worst minute of CPU steal taken as the median across runs so one bad host can't carry a plan.

| # | Plan | Median worst minute | Worst minute seen | Median steal | Runs | Dates |
| --- | --- | --- | --- | --- | --- | --- |
| 1 | [Amazon Lightsail](https://rackcensus.com/vps-providers/lightsail-benchmarks)[Micro, 2 TB transfer](https://rackcensus.com/plans/k4spwg-lightsail-micro-2-tb-transfer-benchmarks) | 88.9% | [89.7%](https://rackcensus.com/benchmarks/o2v9uf-lightsail-micro-2-tb-transfer-us-east-1-benchmark) | 32.0% | 7 runs, 4 locations | Oct 7 to Oct 8 |
| 2 | [OneProvider](https://rackcensus.com/vps-providers/oneprovider-benchmarks)[General Purpose GP1.1-1-10](https://rackcensus.com/plans/zbpld9-oneprovider-general-purpose-gp1-1-1-10-benchmarks) | 24.1% | [24.4%](https://rackcensus.com/benchmarks/m17vxv-oneprovider-general-purpose-gp1-1-1-10-81-benchmark) | 23.0% | 3 runs, 3 locations | Oct 9 to Oct 12 |
| 3 | [DigitalOcean](https://rackcensus.com/vps-providers/digitalocean-benchmarks)[Basic 1 vCPU / 1 GB](https://rackcensus.com/plans/z6r6vs-digitalocean-basic-1-vcpu-1-gb-benchmarks) | 6.4% | [7.2%](https://rackcensus.com/benchmarks/2q6y0c-digitalocean-basic-1-vcpu-1-gb-nyc3-benchmark) | 0.98% | 4 runs, 3 locations | Oct 7 to Oct 8 |
| 4 | [Linode](https://rackcensus.com/vps-providers/linode-benchmarks)[Nanode 1GB](https://rackcensus.com/plans/cigoca-linode-nanode-1gb-benchmarks) | 3.5% | [61.3%](https://rackcensus.com/benchmarks/e0urt1-linode-nanode-1gb-us-west-benchmark) | 0.93% | 6 runs, 4 locations | Oct 7 to Oct 8 |
| 5 | [UpCloud](https://rackcensus.com/vps-providers/upcloud-benchmarks)[Premium 1 vCPU / 1 GB](https://rackcensus.com/plans/9tjgtd-upcloud-premium-1-vcpu-1-gb-benchmarks) | 0.51% | [0.76%](https://rackcensus.com/benchmarks/djcoyk-upcloud-premium-1-vcpu-1-gb-de-fra1-benchmark) | 0.06% | 3 runs, 3 locations | Oct 7 to Oct 10 |
| 6 | [DigitalOcean](https://rackcensus.com/vps-providers/digitalocean-benchmarks)[Basic AMD 1 vCPU / 1 GB](https://rackcensus.com/plans/a382v9-digitalocean-basic-amd-1-vcpu-1-gb-benchmarks) | 0.18% | [7.2%](https://rackcensus.com/benchmarks/odjloi-digitalocean-basic-amd-1-vcpu-1-gb-fra1-benchmark) | 0.03% | 4 runs, 4 locations | Oct 9 to Oct 12 |
| 7 | [UpCloud](https://rackcensus.com/vps-providers/upcloud-benchmarks)[Starter 1 vCPU / 1 GB](https://rackcensus.com/plans/wn4ded-upcloud-starter-1-vcpu-1-gb-benchmarks) | 0.12% | [0.15%](https://rackcensus.com/benchmarks/elq8tn-upcloud-starter-1-vcpu-1-gb-nl-ams1-benchmark) | 0.02% | 6 runs, 4 locations | Oct 7 to Oct 8 |
| 8 | [Scaleway](https://rackcensus.com/vps-providers/scaleway-benchmarks)[PRO2-XXS](https://rackcensus.com/plans/stfucw-scaleway-pro2-xxs-benchmarks) | 0.09% | [0.09%](https://rackcensus.com/benchmarks/iezap6-scaleway-pro2-xxs-nl-ams-1-benchmark) | 0.02% | 3 runs, 3 locations | Oct 8 |
| 9 | [Vultr](https://rackcensus.com/vps-providers/vultr-benchmarks)[Regular Performance Intel 1 vCPU / 1 GB](https://rackcensus.com/plans/x7v9v6-vultr-regular-performance-intel-1-vcpu-1-gb-benchmarks) | 0.04% | [0.1%](https://rackcensus.com/benchmarks/y8l1n6-vultr-regular-performance-intel-1-vcpu-1-gb-ams-benchmark) | 0.0% | 6 runs, 4 locations | Oct 7 to Oct 8 |
| 10 | [Linode](https://rackcensus.com/vps-providers/linode-benchmarks)[Linode 4GB](https://rackcensus.com/plans/jqz5x2-linode-linode-4gb-benchmarks) | 0.04% | [6.3%](https://rackcensus.com/benchmarks/73miwf-linode-linode-4gb-us-east-benchmark) | 0.01% | 3 runs, 3 locations | Oct 10 to Oct 12 |
| 11 | [Vultr](https://rackcensus.com/vps-providers/vultr-benchmarks)[High Performance AMD 1 vCPU / 1 GB](https://rackcensus.com/plans/nn8mia-vultr-high-performance-amd-1-vcpu-1-gb-benchmarks) | 0.03% | [0.04%](https://rackcensus.com/benchmarks/h38o0w-vultr-high-performance-amd-1-vcpu-1-gb-ams-benchmark) | 0.02% | 3 runs, 3 locations | Oct 8 to Oct 10 |

Left out because the host never reported any steal, which usually means it isn't exposed to the server rather than that none happened: [Hetzner CPX22](https://rackcensus.com/plans/sa2c28-hetzner-cpx22-benchmarks).

## How plans get on this list

- A plan needs at least 3 completed runs from the last 90 days, spread across at least 2 locations or 2 different days, so one bad host or one bad afternoon can't rank it.
- Plans are only compared with plans of the same CPU class. Shared CPU plans are never ranked against dedicated ones.
- Every completed run counts. We don't drop bad runs or rerun a plan until it looks better.
- Steal is measured once a second for the whole run, counting only the seconds where the server was asking for at least half its CPU. Steal on an idle server doesn't tell you much.
- A run only counts if it has at least 5 minutes of that kind of load.
- The worst minute is the worst 60-second average in a run. Single-second spikes are mostly noise, so they aren't used.