kafka

我们欢迎大家自由的提问，包括但不限于任何业务需求，方案设计和workflow各种使用细节和实现原理。
但我也建议大家采用一种smart的提问方式，可以参考XY Problem问题：https://xyproblem.info/
XY Problem倡导大家提问时，一定尽可能的描述自己的原始需求，也就是说清楚一件事：你到底想干啥。而不是描述你试图使用的方案，然后让我们帮你解决方案里的问题。因为你采用的方案可能一开始就是错误的。而我们只能一直猜测，你到底想解决什么问题。沟通效率非常低。
例如这个issue（提问者是我们的好朋友，workflow还没开源就在用了。不用怕伤害他：）
sogou/workflow#658
就是一个典型的例子，用户问如何通过派生WFServer的new_connection函数获得proxy

I have noticed when ingesting backlog(older timestamped data) that the "Messages per minute" line graph and "sources" data do not line up.

The Messages per minute appear to be correct for the ingest rate, but the sources breakdown below it only show messages for each type from within the time window via timestamp. This means in the last hour if you've ingested logs from 2 days ago, the data is

我们在使用过程出现了一个问题，就是当我们使用rabbitmq作为队列，并且在使用过程中中断队列，然后重新启动，有一定的概率导致无限的重连。

测试上是部署k8s环境，有两个pod，rabbitmq掉线重连后，其中一个pod正常了，但是另外一个pod一直都在尝试重连
使用代码也比较简单

services.AddCap(option =>
        {
            // 消息存储
            option.UseEntityFramework<FrontDbContext>(o => o.Schema = "FRON

Currently, the COLLECT_SET is only available for windowed stream aggregations, whereas COLLECT_LIST is available for table and stream aggregations.

The reason for this discrepancy is that table and stream aggregations are computed differently in the runtime, and in particular table aggregations need to define a "subtractor" for how to remove a record from the aggregation. (Both table and st

Por Sentry Kafka client (arroyo) tests to Redpanda ducktape tests suite.

https://github.com/getsentry/arroyo/tree/main/tests

This comment says that the message ID is optional,
but for SQL transport it is a mandatory attribute,
in turn it causes misunderstanding?

Is it possible to fix it or did I get something wrong?

https://github.com/ThreeDotsLabs/watermill/blob/b9928e750ba673cf93d442db88efc04706f67388/message/message.go#L20
https://github.com/ThreeDotsLabs/watermill/blob/b9928e750ba673cf93d442db88efc04706f6