Quota configuration helps to achieve fairness and budgeting in service usage.
The metric based quota configuration works this way: - The service configuration defines a set of metrics. - For API calls, the quota.metric_rules maps methods to metrics with corresponding costs. - The quota.limits defines limits on the metrics, which will be used for quota checks at runtime.
An example quota configuration in yaml format:
quota: limits:
- name: apiWriteQpsPerProject
metric: library.googleapis.com/write_calls
unit: "1/min/{project}" # rate limit for consumer projects
values:
STANDARD: 10000
(The metric rules bind all methods to the read_calls metric,
except for the UpdateBook and DeleteBook methods. These two methods
are mapped to the write_calls metric, with the UpdateBook method
consuming at twice rate as the DeleteBook method.)
metricRules:
- selector: "*"
metricCosts:
library.googleapis.com/read_calls: 1
- selector: google.example.library.v1.LibraryService.UpdateBook
metricCosts:
library.googleapis.com/write_calls: 2
- selector: google.example.library.v1.LibraryService.DeleteBook
metricCosts:
library.googleapis.com/write_calls: 1
Corresponding Metric definition:
metrics:
- name: library.googleapis.com/read_calls
displayName: Read requests
metricKind: DELTA
valueType: INT64
- name: library.googleapis.com/write_calls
displayName: Write requests
metricKind: DELTA
valueType: INT64
JSON representation |
---|
{ "limits": [ { object ( |
Fields | |
---|---|
limits[] |
List of QuotaLimit definitions for the service. |
metricRules[] |
List of MetricRule definitions, each one mapping a selected method to one or more metrics. |
QuotaLimit
QuotaLimit
defines a specific limit that applies over a specified duration for a limit type. There can be at most one limit for a duration and limit type combination defined within a QuotaGroup
.
JSON representation |
---|
{ "name": string, "description": string, "defaultLimit": string, "maxLimit": string, "freeTier": string, "duration": string, "metric": string, "unit": string, "values": { string: string, ... }, "displayName": string } |
Fields | |
---|---|
name |
Name of the quota limit. The name must be provided, and it must be unique within the service. The name can only include alphanumeric characters as well as '-'. The maximum length of the limit name is 64 characters. |
description |
Optional. User-visible, extended description for this quota limit. Should be used only when more context is needed to understand this limit than provided by the limit's display name (see: |
defaultLimit |
Default number of tokens that can be consumed during the specified duration. This is the number of tokens assigned when a client application developer activates the service for his/her project. Specifying a value of 0 will block all requests. This can be used if you are provisioning quota to selected consumers and blocking others. Similarly, a value of -1 will indicate an unlimited quota. No other negative values are allowed. Used by group-based quotas only. |
maxLimit |
Maximum number of tokens that can be consumed during the specified duration. Client application developers can override the default limit up to this maximum. If specified, this value cannot be set to a value less than the default limit. If not specified, it is set to the default limit. To allow clients to apply overrides with no upper bound, set this to -1, indicating unlimited maximum quota. Used by group-based quotas only. |
freeTier |
Free tier value displayed in the Developers Console for this limit. The free tier is the number of tokens that will be subtracted from the billed amount when billing is enabled. This field can only be set on a limit with duration "1d", in a billable group; it is invalid on any other limit. If this field is not set, it defaults to 0, indicating that there is no free tier for this service. Used by group-based quotas only. |
duration |
Duration of this limit in textual notation. Must be "100s" or "1d". Used by group-based quotas only. |
metric |
The name of the metric this quota limit applies to. The quota limits with the same metric will be checked together during runtime. The metric must be defined within the service config. |
unit |
Specify the unit of the quota limit. It uses the same syntax as [Metric.unit][]. The supported unit kinds are determined by the quota backend system. Here are some examples: * "1/min/{project}" for quota per minute per project. Note: the order of unit components is insignificant. The "1" at the beginning is required to follow the metric unit syntax. |
values |
Tiered limit values. You must specify this as a key:value pair, with an integer value that is the maximum number of requests allowed for the specified unit. Currently only STANDARD is supported. An object containing a list of |
displayName |
User-visible display name for this limit. Optional. If not set, the UI will provide a default display name based on the quota configuration. This field can be used to override the default display name generated from the configuration. |