admin管理员组

文章数量:1124536

I have one pod on GKE with Istio, which streams data using gRPC. I am streaming a lot of avro records to a client. After some time, usually max 30 minutes, the connection is reset with error: http2.remote_reset

While debugging the client I see that there is OUTBOUND GO_AWAY: lastStreamId=2147483647 errorCode=2

and then: Sending GOAWAY failed: lastStreamId '2147483647' errorCode '2'

Every time it is the same lastStreamId.

I've set max_outbound_frames: 1000000 in EnvoyFilter I have following keepalive settings for grpc server:
keep-alive-time: 60s keep-alive-timeout: 20s

The grpc server is written in Java with Spring Boot, using net.devh library.

Is the amount of data the problem here? Is there a way to make it work from server side? There will be different consumers for this grpc service in the future, but right now I cannot make it work even with my sample client.

Edit: One point to mention, I have the same grpc service running onprem, without Istio, and the issue doesn't appear there.

本文标签： google cloud platformgRPC server receiving http2remotereset when sending many messagesStack Overflow

版权声明：本文标题：google cloud platform - gRPC server receiving http2.remote_reset when sending many messages - Stack Overflow 内容由网友自发贡献，该文观点仅代表作者本人， 转载请联系作者并注明出处：http://www.betaflare.com/web/1736635892a1945879.html， 本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如发现本站有涉嫌抄袭侵权/违法违规的内容，一经查实，本站将立刻删除。