The Minimum Number of Concurrent Online Instances Supports Auto Scaling. Is It OK to Set the Minimum Number of Concurrent Online Instances?
Currently, the recommended maximum number of concurrent online instances is 5. The scaling resource has a certain delay, which slows down the prediction APIs. You are therefore advised to predict and enter possible concurrency specifications. The system reserves resources in advance based on concurrency specifications to avoid peaks.
If you need more specifications, submit a service ticket to Huawei technical support.
Did this article solve your problem?
Thank you for your score!Your feedback would help us improve the website.