Limits
Rate Limit¶
Grader Than Workspaces are rate-limited to 100 requests per second. Exceeding this limit will result in a 429 HTTP error response regardless the underlying service's response. As workspaces on Grader Than are primarily used for academic purposes, the 100 requests per second rate limit should be sufficient for most academic use cases.
What is a rate limit?
A rate limit is a mechanism used by web services to restrict the number of requests a client can make in a given timeframe. It prevents server overload, slow response times, and service crashes by ensuring fair usage.
Why are there rate limits?
Rate limits are used to prevent excessive usage of Grader Than services, which can cause server overload, slow down response times, and even crash our services. By enforcing a rate limit, Grader Than can ensure that our service remains available and responsive for all users.
Without rate limits, some users may consume an excessive amount of server resources, which could impact the performance of our service for other users. Rate limits help to ensure that resources are allocated fairly, and that users are able to access the service in a reliable and consistent manner.
In addition to preventing service overload, rate limits can also help to protect the security of our service and prevent abuse by malicious users. By limiting the number of requests that can be made within a given timeframe, rate limits help to prevent denial-of-service attacks and other types of malicious behavior that could compromise the service.
Overall, rate limits are an important mechanism for maintaining the performance, reliability, and security of Grader Than services, and help to ensure that all users are able to access our services in a fair and consistent manner.
Timeouts¶
The Grader Than system will emit a 408 HTTP error response if any workspace services request does not complete within 30 seconds. This timeout mechanism helps ensure that the system remains responsive and performant, even in the face of slow or unresponsive services.
If you receive a 408 error response, it indicates that the request has timed out and should be retried. It's important to note that this timeout is not configurable by users, as it is set by the Grader Than system to ensure consistent and reliable performance across all workspaces.
What is a timeout
A timeout is a mechanism used to limit the amount of time the Grader Than system waits for a response from a custom service before giving up and assuming that the request has failed. Timeouts are commonly used to ensure that requests do not hang indefinitely and to prevent Grader Than server resources from being tied up by slow or unresponsive custom services.