Skip to content
Branch: master
Find file Copy path
Fetching contributors…
Cannot retrieve contributors at this time
98 lines (86 sloc) 4.07 KB
syntax = "proto3";
option go_package = "ratelimit";
package pb.lyft.ratelimit;
service RateLimitService {
// Determine whether rate limiting should take place.
rpc ShouldRateLimit (RateLimitRequest) returns (RateLimitResponse) {}
// Main message for a rate limit request. The rate limit service is designed to be fully generic
// in the sense that it can operate on arbitrary hierarchical key/value pairs. The loaded
// configuration will parse the request and find the most specific limit to apply. In addition,
// a RateLimitRequest can contain multiple "descriptors" to limit on. When multiple descriptors
// are provided, the server will limit on *ALL* of them and return an OVER_LIMIT response if any
// of them are over limit. This enables more complex application level rate limiting scenarios
// if desired.
message RateLimitRequest {
// All rate limit requests must specify a domain. This enables the configuration to be per
// application without fear of overlap. E.g., "envoy".
string domain = 1;
// All rate limit requests must specify at least one RateLimitDescriptor. Each descriptor is
// processed by the service (see below). If any of the descriptors are over limit, the entire
// request is considered to be over limit.
repeated RateLimitDescriptor descriptors = 2;
// A RateLimitDescriptor is a list of hierarchical entries that are used by the service to
// determine the final rate limit key and overall allowed limit. Here are some examples of how
// they might be used for the domain "envoy".
// 1) ["authenticated": "false"], ["ip_address": ""]
// What it does: Limits all unauthenticated traffic for the IP address The
// configuration supplies a default limit for the ip_address field. If there is a desire to raise
// the limit for or block it entirely it can be specified directly in the
// configuration.
// 2) ["authenticated": "false"], ["path": "/foo/bar"]
// What it does: Limits all unauthenticated traffic globally for a specific path (or prefix if
// configured that way in the service).
// 3) ["authenticated": "false"], ["path": "/foo/bar"], ["ip_address": ""]
// What it does: Limits unauthenticated traffic to a specific path for a specific IP address.
// Like (1) we can raise/block specific IP addresses if we want with an override configuration.
// 4) ["authenticated": "true"], ["client_id": "foo"]
// What it does: Limits all traffic for an authenticated client "foo"
// 5) ["authenticated": "true"], ["client_id": "foo"], ["path": "/foo/bar"]
// What it does: Limits traffic to a specific path for an authenticated client "foo"
// The idea behind the API is that (1)/(2)/(3) and (4)/(5) can be sent in 1 request if desired.
// This enables building complex application scenarios with a generic backend.
message RateLimitDescriptor {
message Entry {
string key = 1;
string value = 2;
repeated Entry entries = 1;
// Defines an actual rate limit in terms of requests per unit of time and the unit itself.
message RateLimit {
enum Unit {
HOUR = 3;
DAY = 4;
uint32 requests_per_unit = 1;
Unit unit = 2;
// A response from a ShouldRateLimit call.
message RateLimitResponse {
enum Code {
OK = 1;
message DescriptorStatus {
// The response code for an individual descriptor.
Code code = 1;
// The current limit as configured by the server. Useful for debugging, etc.
RateLimit current_limit = 2;
// The limit remaining in the current time unit.
uint32 limit_remaining = 3;
// The overall response code which takes into account all of the descriptors that were passed
// in the RateLimitRequest message.
Code overall_code = 1;
// A list of DescriptorStatus messages which matches the length of the descriptor list passed
// in the RateLimitRequest. This can be used by the caller to determine which individual
// descriptors failed and/or what the currently configured limits are for all of them.
repeated DescriptorStatus statuses = 2;
You can’t perform that action at this time.