Hasura Cloud Data Usage Calculation
How do we calculate data usage?
Hasura Cloud projects are billed based on the amount of data that passes through Hasura Cloud. This data corresponds to the operations that are performed by you, your apps, or your clients.
The data pass-through is calculated by analyzing the logs that are sent from your project. All data transfer from Hasura Cloud to external networks - such as your database server, external webhooks for Actions, Event Triggers, Scheduled Triggers or your Remote Schema - is accounted for in Hasura Cloud billing.
We account for the following types of logs in Hasura Cloud data pass-through:
http-log
: The data usage is calculated by the request and response sizes ofhttp-log
. HTTP request size is dependent on the size of cookies and request headers. Streamlining the response data will help reduce the HTTP response size and corresponding data usage. Client-side errors like validation, query syntax, and rate-limiting errors are not accounted for in billing.ws-server
: The volume of data sent and received through a WebSocket connection is considered when calculating your data usage. WebSocket connections are typically used for subscriptions. You can optimize the amount of data sent via subscriptions to reducews-server
data usage.action-handler-log
: The sizes of the request and response ofaction-handler-log
is taken into account for calculating your data usage.event-trigger-log
: The sizes of the request and response ofevent-trigger-log
is taken into account for calculating your data usage. This log is generated when Hasura triggers an external webhook.scheduled-trigger-log
: The sizes of the request and response of thescheduled-trigger-log
are taken into account for calculating your data usage. The cost incurred by this depends on the frequency of Scheduled Triggers.
These logs are used to calculate your data usage in Hasura Cloud, which is then used for generating monthly invoices. You can find more information about the pricing for data pass-through on our pricing page.