Limits and quotas
This content is not available in your language yet.
There are three kinds of limit. Tier limits and project quotas can be raised; fixed limits are part of how the platform works. See Quotas and tiers for how tiers work and how to check your usage.
Organization limits by tier
Section titled “Organization limits by tier”new |
payment_verified |
id_verified |
|
|---|---|---|---|
| Projects | 2 | 5 | 10 |
| VMs (all projects) | 2 | 10 | 25 |
| vCPUs (all projects) | 2 | 16 | 48 |
| Memory (all projects) | 4 GiB | 32 GiB | 96 GiB |
| Public IPv4 addresses (all projects) | 1 | 3 | 5 |
| Load balancers (all projects) | 1 | 3 | 5 |
| Buckets (all projects) | 3 | 20 | 50 |
| Stored in buckets (all projects) | 10 GiB | 250 GiB | 1,000 GiB |
| Public buckets | no | yes | yes |
| Bandwidth per VM, each direction | 50 Mbit/s | 200 Mbit/s | 1,000 Mbit/s |
Project quotas (defaults)
Section titled “Project quotas (defaults)”| Quota | Default |
|---|---|
| VMs | 5 |
| vCPUs | 8 |
| Memory | 16 GiB |
| Volumes, including boot volumes | 20 |
| Total volume size | 100 GiB |
| Snapshots (one per disk in each VM snapshot) | 20 |
| Snapshot schedules | 10 |
| VPCs, including the default | 5 |
| Subnets per VPC | 20 |
| Security groups, including default groups | 50 |
| Rules per security group | 60 |
| Public IPv4 addresses | 2 |
| Load balancers | 2 |
| Listeners per load balancer | 10 |
| Target groups per load balancer | 10 |
| Targets per load balancer | 50 |
| Buckets | 10 |
| Stored in buckets | 100 GiB |
| S3 keys | 20 |
Fixed limits
Section titled “Fixed limits”| What | Limit |
|---|---|
| Organizations a person can own | 3 |
| Volume size | 1 to 2,048 GiB |
| Data volumes per VM | 16 |
| Snapshot schedules per VM | 5 |
| Snapshot schedule frequency | At most once an hour |
| Snapshot schedule retention | 1 to 100 snapshots, 1 to 365 days |
| Security groups per VM | 1 to 5 |
| SSH keys per VM at creation | 20 |
| cloud-init user data | 32 KiB |
| Public IPs per VM | 1 |
| NAT gateways per VPC | 1 |
| Load balancer nodes | 2 per load balancer, in its VPC |
| Load balancer health check | every 5 to 300 seconds, 2 to 10 checks to change state |
| VPC IPv4 range | /16 to /28, inside 10.0.0.0/8, 172.16.0.0/12 or 192.168.0.0/16 |
| Policy versions | 5 per policy |
| Policy document | 6,144 characters (trust policies 2,048) |
| Role session length | 15 minutes to 12 hours (1 hour when started from another role session) |
| Access key lifetime | 90 days by default, at most 365 days (service accounts may have non-expiring keys) |
| Invitation validity | 7 days |
| Console ticket | 60 seconds to connect |
| Released public IPv4 cool-down | 24 hours before it is handed out again |
| M-Pesa top-up | KES 100 to 150,000 |
| Outbound TCP port 25 | Blocked |
| New outbound connections | 500 a second (bursts to 1,000) per public IPv4 address or VPC IPv6 range |
API rate limits
Section titled “API rate limits”The API limits how fast one client can call it, to keep it responsive for everyone. A client over
the limit gets 429 Too Many Requests; wait and retry with backoff. Signups are limited to a few
per hour from one address.