Resources per replica#
| Limit | Shared compute | Isolated compute (dedicated) |
|---|
| Minimum CPU | 0.125 vCPU (125m) | 0.125 vCPU (125m) |
| Maximum CPU | 4 vCPU | The machine type's vCPU |
| Minimum memory | 256 MiB | 256 MiB |
| Maximum memory | 8 GiB | The machine type's memory |
| Memory per vCPU | 1–4 GiB per vCPU | No ratio rule |
| Machine sizes | — | 2 or 4 vCPU today. Larger sizes are coming. |
Builds and deploys#
| Limit | Value |
|---|
| Container port | 1 – 65535. Optional, but required to expose a deploy. |
| Replicas | At least 1 |
| Environments per deploy | 1, fixed when the deploy is created |
| Project of a build | Fixed when the build is created |
Secrets#
| Limit | Value |
|---|
| Name characters | Letters, digits, . _ = -. No spaces, slashes or non-ASCII characters. |
| Name length | Up to 253 characters |
| Scope | One environment |
| Reading values back | Not possible. Values are write-only. |
Domains and routes#
| Limit | Value |
|---|
| Time to verify a claim | 72 hours, then the claim is marked Failed |
Wildcard hostnames (*.example.com) | Not supported. Rejected when you attach a route. |
| Deploys sharing one hostname | Allowed by path prefix, if they are all in the same environment and location |
| Locations per hostname | One. Use a hostname per location. |
| Apex domains | Need ALIAS/ANAME/CNAME flattening, or A records on request |
Logs and metrics#
| Limit | Value |
|---|
| Lines kept in the logs viewer | 5,000 (the oldest drop off) |
| Metrics ranges (deploy, pod, node) | 5m, 15m, 1h, 6h, 24h |
| Metrics ranges (organisation) | 1h, 6h, 24h, 7d |
| Node metrics | Dedicated nodes only |
Not yet published#
If you're planning a large workload, contact info@devlyft.io first.