opensergo-control-plane icon indicating copy to clipboard operation
opensergo-control-plane copied to clipboard

Universal cloud-native microservice governance control plane (微服务治理控制面)

Results 37 opensergo-control-plane issues
Sort by recently updated
recently updated
newest added

Signed-off-by: Jiangnan Jia ### Describe what this PR does / why we need it Support Helm Chart deploy ### Does this pull request fix one issue? Fixes #40 ### Describe...

area/build-release

## Issue Description Type: *feature request* ### Describe what feature you want Add Helm Chart deployment suite for control plane

kind/feature
area/build-release

Signed-off-by: Jiangnan Jia ### Describe what this PR does / why we need it Add arg of GOPROXY in Dockerfile, so we can change the goproxy when build image.

area/build-release

In Istio environment, we need an extra watcher to watch OpenSergo CRDs actively, so we can know the change of OpenSergo CRDs such as TrafficRouter and etc. When the OpenSergo...

kind/feature

Add user-define configuration feature and retry feature for pushing config. The main idea is to add retry ability for `stream.SendMsg`. From the general perspective, adopt third-party lib `go-retry` for this...

## Issue Description Type: *feature request* ### Describe what feature you want The Current base-image in Dockerfile is the `** alpine **`, and this image lacks some useful tools for...

kind/enhancement
help wanted

Add fundamental logger mechanism. --- 目前 OpenSergo 控制面的日志机制非常简单,复用了 Sentinel Go 的日志模块。我们需要设计一个通用的 logger 机制与基本实现,并支持任意扩展。

kind/feature
area/logging

在 OpenSergo universal transport service protocol (OUTS) 中,控制面通过下行链路 (server data-push) 向 OpenSergo client 推送订阅的数据,client 收到数据后会返回 response。这个过程需要一个超时机制,当 server 端 push data 一段时间后仍未收到该 push 请求的回复,则需要有一定的异常处理机制(如重试)。欢迎社区一起来设计与实现。 --- Add timeout mechanism for config-push transport...

good first issue
kind/feature

Add debounce mechanism for CRD event handling in CRDWatcher. --- CRDWatcher 处理 CRD 变更事件时需要有 debounce 机制,避免同个 subscribeTarget(a, n, k) 短时间内多次变更导致频繁触发 push。在全量推送模型中,可以有一个简单的蓄水池机制,比如累积 100ms 的变更再合并推送。 debounce 的范围以及 timeout 需要做到可配置。欢迎社区针对这个机制进行讨论。

good first issue
kind/feature
area/crd-watcher

Support sub-push CRD with LabelKVs scope --- 当前 控制面 与 SDK 的 订阅/推送 机制中,SubscribeTarget 中的 SubscribeLabelKVs 不生效,因此仅支持 Namespace-App-Kind 订阅组范围内的数据推送,即当 同一个 Namespace-App-Kind 范围内中 任意一个 CRD 资源对象发生变动时,都会触发 推送动作,推送的数据是 整个 Namespace-App-Kind 订阅组范围内的 CRD...

kind/feature