Ask HN: 如何在负载高峰时保持语音 AI 的低延迟？
1 分•作者: didro•7 个月前
大家好，
我正在做一项研究，有个问题想请教一下。当语音 AI 助手的流量激增时，你们会怎么处理？据我所知，Kubernetes 的原生自动伸缩器（HPA）通常无法及时响应这种突发流量，导致延迟过高。如果方便的话，我很乐意了解一下你们的经验。
谢谢！
查看原文
Hey guys,<p>I'm doing a research and have a question. What do you do when traffic to voice AI agents spikes? As far as I know, native autoscaler of Kubernetes (HPA) doesn't catch up quite often with that - resulting into a prohibitively high latency. Would be glad to know your experience if you don't mind.<p>Thanks!