We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? # to your account
在k8s集群中,限制容器系统资源,在高并发下,客户端执行rpc调用时会出现全部请求都卡住无响应的情况,grpc服务端未收到调用请求,客户端也不会超时,没有异常信息,不确定是不是死锁。配置除了注册中心地址,其他基本使用默认配置。不接入polaris注册中心时不会出现。服务端是用go提供的grpc服务。 下图是可能的异常堆栈:
环境版本: 应用基础镜像:openjdk:8-jdk-alpine polaris-java版本: 1.10.1, 1.10.4 grpc版本:1.47.0,1.52.1
The text was updated successfully, but these errors were encountered:
好的,这周会将改问题解决
Sorry, something went wrong.
周末修复了下可能存在的问题点,还在本地复现验证中
下周我们会对 grpc-java-polaris 进行整体压测看看
好的
更新:基本解决,目前定位是由于处理队列(服务端和客户端)设置为固定值,且策略为丢弃导致
请问这个问题是什么原因导致的? 是这样的,我没有使用本工程,但是目前我在使用grpc时也遇到了本issue提到的问题,但我尚未定位问题的原因。 所以想了解该项目遇到的问题是否能帮助我解决问题,感激不尽。
问题描述: rpc:Grpc 环境:客户端和服务端都是MacOS proto:rpc publish(Message message) {return Response} 过程:
其他:
chuntaojun
No branches or pull requests
在k8s集群中,限制容器系统资源,在高并发下,客户端执行rpc调用时会出现全部请求都卡住无响应的情况,grpc服务端未收到调用请求,客户端也不会超时,没有异常信息,不确定是不是死锁。配置除了注册中心地址,其他基本使用默认配置。不接入polaris注册中心时不会出现。服务端是用go提供的grpc服务。
下图是可能的异常堆栈:
环境版本:
应用基础镜像:openjdk:8-jdk-alpine
polaris-java版本: 1.10.1, 1.10.4
grpc版本:1.47.0,1.52.1
The text was updated successfully, but these errors were encountered: