This topic describes the resource limits in Function Compute.
Resource limits for services
|Functions per service||50|
|Triggers per function||10|
Resource limits for function execution
|Item||Maximum limit (flexible instance)||Maximum limit (performance instance)|
|Temporary disk space (the /tmp directory)||512 MB||10 GB|
|Total processes and threads||1024||1024|
|Memory required to execute a function||3 GB||16 GB|
|Duration for which a function is executed||900s||7200s|
|Duration for which an initializer is executed||300s||300s|
|Response payload size of synchronous invocation||6 MB||6 MB|
|Request payload size of asynchronous invocation||128 KB||128 KB|
|Deployment package size (ZIP or JAR file)||100 MB||500 MB|
|Source code size||500 MB||10 GB|
Resource limits per region per account
|Total size of uploaded deployment packages||100 GB|
Note If you need more pay-as-you-go instances, submit a ticket.
|Single instance concurrency||1~100|
Limits on accessing other Alibaba Cloud services or resources
If the feature of accessing resources in a virtual private cloud (VPC) has been enabled for Function Compute in your region, the following networking limits apply to your calls from Function Compute to other Alibaba Cloud services or resources.
- You cannot use the internal IP address of an Elastic Compute Service (ECS) instance in a classic network to access its resources. These resources include web services and file systems. You must use a public IP address to access these resources or migrate these resources to a VPC.
- You cannot access an ApsaraDB for RDS instance in a classic network by using its internal IP address. You must use its public IP address or migrate it to a VPC.
- You cannot access an Alibaba Cloud service by using its internal endpoint. You must use a VPC endpoint or public endpoint provided by this service.