You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Our API’s are SLA bound returning responses within 2 secs with multiple integrations involved, to avoid misuse and slowing down downstream systems with too many requests in a given time period just delay the response.
We want to delay responses based on a criteria like requests from specific IP address. Although we want to honor each requests within 2 secs, at the same time if we want to minimize the misuse from a specific integration partner/customer, we don’t want to reject the request, rather delay the response after a defined threshold.
If we can apply Latency policy in such a way like if in a given second we receive more than 50 requests from a particular IP address, apply Latency of 10 secs on subsequent requests for next one minute for all requests from the same IP address.
This way we want to ensure we can minimize the misuse else it may effect the SLA for all customers/integration partners, protect downstream systems from overload so those systems can still perform within defined SLA and then also make sure our customer/integration partner is getting the response although little delayed (if its beyond threshold) but not rejecting the request by applying Spike Arrest or Rate Limit.
The text was updated successfully, but these errors were encountered:
Our API’s are SLA bound returning responses within 2 secs with multiple integrations involved, to avoid misuse and slowing down downstream systems with too many requests in a given time period just delay the response.
We want to delay responses based on a criteria like requests from specific IP address. Although we want to honor each requests within 2 secs, at the same time if we want to minimize the misuse from a specific integration partner/customer, we don’t want to reject the request, rather delay the response after a defined threshold.
If we can apply Latency policy in such a way like if in a given second we receive more than 50 requests from a particular IP address, apply Latency of 10 secs on subsequent requests for next one minute for all requests from the same IP address.
This way we want to ensure we can minimize the misuse else it may effect the SLA for all customers/integration partners, protect downstream systems from overload so those systems can still perform within defined SLA and then also make sure our customer/integration partner is getting the response although little delayed (if its beyond threshold) but not rejecting the request by applying Spike Arrest or Rate Limit.
The text was updated successfully, but these errors were encountered: