CloudS3n

Results 11 comments of CloudS3n

I think the logic behind the dependent node design should be that for it to be considered successful, every time slice needs successful results. For periods like this month or...

Currently, I can only set the baseUrl to the context path. ``` server: port: 8080 servlet: context-path: /datakits/ai/web spring: main: web-application-type: servlet ai: mcp: server: name: webmvc-mcp-server version: 1.0.0 type:...

Maybe we need a cache queue for the RECOVER_SERIAL_WAIT command? If the command execution finds that the previous workflow has not been updated to a completed state in time, it...

Currently, I resolved this issue by retrying the RECOVER_SERIAL_WAIT command when the previous workflow was not fully updated, and it works well.

> Right now the serial wait implementation is really unstable, there are a lot of case will cause it doesn't work well. e.g. > > 1. Concurrent trigger will cause...

@Gallardot Yes, I’ve already applied patch #15270. But as I mentioned earlier, the state change is always a process. Opening a new transaction only speeds up the state update, but...

@Gallardot Here’s another example for `saveSerialProcess`: In concurrent scenarios (multiple parent workflows with the same scheduling cycle reference the same sub-workflow), multiple start_workflow commands will be consumed by different masters...

This feature is great, thanks for your contribution🎉

Obviously, the maintainers of this project do not have the energy to maintain it.