Quotas
This document describes the quota limits for Google Cloud Functions.
Quotas for Google Cloud Functions encompass 3 areas:
Resource Limits
These affect the total amount of resources your functions can consume.
Time Limits
These affect how long things can run.
Rate Limits
These affect the rate at which you can call the Cloud Functions API to manage your functions.
The different types of limits are described in more detail below. Differences between limits for Cloud Functions (1st gen) and Cloud Functions (2nd gen) are noted where applicable.
Resource Limits
Resource limits affect the total amount of resources your functions can consume. The regional scope is per project, and each project maintains its own limits.
Quota | Description | Limit (1st gen) | Limit (2nd gen) | Can be increased | Scope |
---|---|---|---|---|---|
Number of functions | The total number of functions that can be deployed per region | 1,000 | 1,000 minus the number of Cloud Run services deployed | No | per region |
Max deployment size | The maximum size of a single function deployment | 100MB (compressed) for sources. 500MB (uncompressed) for sources plus modules. |
N/A | No | per function |
Max uncompressed HTTP request size | Data sent to HTTP Functions in an HTTP request | 10MB | 32MB | No | per invocation |
Max uncompressed HTTP response size | Data sent from HTTP functions in an HTTP response | 10MB | 10MB for streaming responses. 32MB for non-streaming responses. |
No | per invocation |
Max event size for event-driven functions | Data sent in events to background functions | 10MB | 512KB for Eventarc events. 10MB for legacy events. |
No | per event |
Max function memory | Amount of memory each function instance can use | 8GiB | 32GiB | No | per function |
Max project memory | Amount of memory, in By, that a project can use. It is measured by the total sum of user-requested memory across function instances over a 1 minute period. | Depends on selected region. This limit might be greater in high-capacity regions or lower in recently opened regions. | N/A | Yes | per project and region |
Max project CPU | Amount of CPU, in milli vCPU, that a project can use. It is measured by the total sum of user-requested CPU across function instances over a 1 minute period. | Depends on selected region. This limit might be greater in high-capacity regions or lower in recently opened regions. | N/A | Yes | per project and region |
Time Limits
Quota | Description | Limit (1st gen) | Limit (2nd gen) | Can be increased | Scope |
---|---|---|---|---|---|
Max function duration | The maximum amount of time a function can run before being forcibly terminated | 540 seconds | 60 minutes for HTTP functions. 9 minutes for event-driven functions. |
No | per invocation |
Rate Limits
Quota | Description | Limit (1st gen) | Limit (2nd gen) | Can be increased | Scope |
---|---|---|---|---|---|
API calls (READ) | Calls to describe or list functions via the Cloud Functions API | 5000 per 100 seconds | 1200 per 60 seconds | Only for 1st gen | per project (1st gen) per region (2nd gen) |
API calls (WRITE) | Calls to deploy or delete functions via the Cloud Functions API | 80 per 100 seconds | 60 per 60 seconds | No 1 | per project (1st gen) per region (2nd gen) |
API calls (CALL) | Calls to the "call" API | 16 per 100 seconds | N/A | No 2 | per project |
Scalability
Cloud Functions invoked by HTTP scale up quickly to handle incoming traffic, while background functions scale more gradually. A function's ability to scale up is dictated by a few factors, including:
- The amount of time it takes for a function's execution to complete (short-running functions can generally scale up to handle more concurrent requests).
- The amount of time it takes for a function to initialize on cold start.
- Your function's error rate.
Transient factors, such as regional load and data center capacity.
Your configuration as defined by minimum instances, maximum instances, and concurrency (concurrency is 2nd gen only).
Additional quotas for background functions
Quota | Description | Limit | Can be increased | Scope | Product version |
---|---|---|---|---|---|
Max concurrent invocations | The maximum concurrent invocations of a single function Example: if handling each event takes 100 seconds, the invocation rate will be limited to 30 per second on average |
3,000 | Yes | per function | 1st gen only |
Max invocation rate | The maximum rate of events being handled by a single function Example: if handling an event takes 100ms, the invocation rate will be limited to 1000 per second even if only 100 requests, on average, are handled in parallel |
1000 per second | No | per function | 1st gen only |
Max concurrent event data size | The maximum total size of incoming events to concurrent invocations of
a single function Example: if events are of size 1MB and processing them takes 10 seconds, the average rate will be 1 event per second, because the 11th event will not be processed until processing one of the first 10 events finishes |
10MB | No | per function | 1st gen and 2nd gen |
Max throughput of incoming events | The maximum throughput of incoming events to a single function Example: if events are of size 1MB, then the invocation rate can be maximum 10 per second, even if functions finish within 100ms |
10MB per second | No | per function | 1st gen and 2nd gen |
When you reach a quota limit
When a function consumes all of an allocated resource, the resource becomes unavailable until the quota is refreshed or increased. This may mean that your function and all other functions in the same project will not work until then. A function returns an HTTP 500 error code when one of the resources is over quota and the function cannot execute.
To increase quotas above the defaults listed here, go to the Cloud Functions Quotas Page, select the quotas you want to modify, click EDIT QUOTAS, supply your user information if prompted, and enter the new quota limit for each quota you selected.