Shared web server performance and web service scalability
To enhance shared web server performance and scalability, you can enable listener processes to run in Low Latency mode on dedicated execution workers.
Enable Web Services Server listener processes to execute mainly in memory by selecting the Low Latency option for them.
If your web service is deployed to a Runtime in a runtime cloud, dedicate a reserved resource called an execution worker to processing web service requests for your account. Performance with an execution worker enabled should rival that of an ordinary Runtime or Runtime cluster.
Shared web server traffic control
A runtime cloud owner can set multiple properties per runtime cloud cluster or attach Runtime to control shared web server traffic on the Attachment Quotas tab on the Cloud Management page.
The following properties can be set by the runtime cloud owner per runtime cloud cluster or attached Runtime to control shared web server traffic:
- maximum number of requests per second
- maximum input size for requests
- maximum number of concurrent executions
For a runtime cloud on which execution workers are enabled, the owner can additionally set the following properties per runtime cloud cluster or attached Runtime:
- maximum number of concurrent executions per execution worker
- maximum process execution time
- maximum number of extra queued processes when the maximum number of processes is running
- maximum time that a queued job can wait to be executed
These properties are set on the Attachment Quotas tab in the Cloud Management page (Manage > Cloud Management).