Resource limits
¶ Cluster limits
Clusters limit is set per account. The default value is set to 2 clusters. If you need more clusters, contact our Helpdesk for the limit extension.
¶ Worker limits
The overall limit is 10 workers per cluster. If you need more workers, contact our Helpdesk for a worker amount increase.
The number of workers is currently limited to 10 due to the specifics of our current packet routing mode. In the following versions, IPVS implementation is planned to withdraw worker limitations and allow for bigger clusters without system overload.
¶ Private cloud limits
The maximum number of workers accessible in the private cloud depends on your available private cloud resources and your workers configuration.
¶ Public cloud limits
The maximum number of available workers depends on resource availability according to your current plan, workers in pending (being created) state count, and Autoscaler resource reservations.
¶ Global resource limits
are set per account for all VPSs and applied throughout the cloud server bound to a worker.
- IPs: 10 by default
- vCPUs: 32 by default
Contact our support team to change the general VPS limits for your account.
¶ Resource limits
The Worker Pool settings page widget displays used and available resources along with the maximum number of workers the worker pool can contain.
See the example below for 4 workers, 9 vCPUs and 9979 IPs available to Kubernetes user in Websa panel.
¶ Autoscaler commitments
Resource reservations are made according to the maximum values you set when configuring your autoscaler. To ensure your worker pool’s ability to scale up, we make resource commitments for those workers that don’t exist yet but should be created when scaling up.
You won’t pay for the reservations, but we will deduct the commitment amount from the resources available to you. Read more about our Autoscaler.
Workers commitment calculation example:
Autoscaler is enabled, the minimum workers amount is set to 1, and the maximum amount is set to 4 workers. In this case, reservations will be made for 3 workers.
¶ Available workers calculation example
Client’s limit is 32 VCPUs
Plan VCPU value for current worker pool: 2 VCPU
There are 2 pending workers within the same plan (of 2 VCPU)
Client’s cloud servers currently utilize 16 VCPUs
An Autoscaler is on for another worker pool of the same cluster, the pool uses the same plan and currently contain 2 workers, Autoscaler minimum value is 2 workers and the maximum is 4 workers.
Therefore, we reserve 4 more VCPUs (2 VCPUs per worker).
Available workers count = (32 (global limit) - 2x2 (pending workers) - 16 (used VCPUs) - 2x2 (resource reservations)) / 2 (current pool plan value) = 4.
Result:
In this case, the client can create 4 workers at this point.