Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[CT-1048] As part of proto integration convert NodeFinished class and run_results to use proto definitions #5663

Closed
gshank opened this issue Aug 17, 2022 · 6 comments
Assignees
Labels

Comments

@gshank
Copy link
Contributor

gshank commented Aug 17, 2022

The proto events will need to be generated outside of our normal class hierarchy. We need to establish best practices and methods for doing this easily, since we'll need node output for all of our node types. In addition we need proto definitions for RunResults.

As part of this effort we will collect requirements from other stakeholders to use when choosing which information to include and to plan additional events.

@github-actions github-actions bot changed the title As part of proto integration convert NodeFinished class and run_results to use proto definitions [CT-1048] As part of proto integration convert NodeFinished class and run_results to use proto definitions Aug 17, 2022
@gshank gshank added the logging label Aug 17, 2022
@leahwicz
Copy link
Contributor

@gshank could you edit this ticket to better describe what we talked about

@gshank
Copy link
Contributor Author

gshank commented Sep 28, 2022

Will identify a list of existing events that would be most useful for updating the progress of execution and parsing, and if there are gaps, create new events. In addition will create logging events that will contain some of the information that is currently retrieved from run_result.json.

@gshank gshank self-assigned this Sep 28, 2022
@gshank
Copy link
Contributor Author

gshank commented Sep 28, 2022

A related part of this is identifying ways to generate proto definitions from Python classes, because creating a set of parallel classes where every time we add a new attribute it must be added in two places is suboptimal. This might look like adding proto integer identifiers to attributes, like pure_protobuf.

@jtcohen6
Copy link
Contributor

jtcohen6 commented Oct 3, 2022

Split the spike/investigation into a new ticket: #5987

Let's keep this ticket focused on:

  • the info from NodeFinished that can support the end-to-end Model Timing MVP with Metadata
  • identifying gaps in the logs, vis-a-vis what Metadata/IDE need — e.g. decorating more events with NodeInfo

@gshank
Copy link
Contributor Author

gshank commented Oct 7, 2022

@jtcohen6 Question: We have 22 logging events that include node_info right now. But a lot of them are variants such as "PrintModelResultLine" and "PrintModelErrorResultLine". These events match the stdout lines that we produce in a terminal. But a consumer of the logs is going to have to subscribe to multiple events for every one of the node types in order to capture both success and failure.

I'm wondering if it might not work better to have a single "ModelResult", which contains a success attribute, and can produce either the success or error message.

@emmyoop
Copy link
Member

emmyoop commented Dec 1, 2022

This has been done as part other tickets.

@emmyoop emmyoop closed this as completed Dec 1, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

Successfully merging a pull request may close this issue.

4 participants