-
Notifications
You must be signed in to change notification settings - Fork 1
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
cross-back-end process graph processing #150
Comments
@JeroenVerstraelen can you point me to the details on this feasibility study? |
decouple "load_collection" detection from subjob creation loop
… yielding Introduce ProcessGraphSplitterInterface, with first LoadCollectionGraphSplitter implementation based on existing simple graph splitting logic
decouple "load_collection" detection from subjob creation loop
… yielding Introduce ProcessGraphSplitterInterface, with first LoadCollectionGraphSplitter implementation based on existing simple graph splitting logic
also make graph walking more deterministic (e.g. to simplify test asserts)
Instead of dictionary Also move type coercion logic to _FrozenNode __init__ to eliminate boilerplate code
also add consistency check of _GraphViewer node map
…orBatchJobs._create_crossbackend_job
decouple "load_collection" detection from subjob creation loop
… yielding Introduce ProcessGraphSplitterInterface, with first LoadCollectionGraphSplitter implementation based on existing simple graph splitting logic
also make graph walking more deterministic (e.g. to simplify test asserts)
Instead of dictionary Also move type coercion logic to _FrozenNode __init__ to eliminate boilerplate code
also add consistency check of _GraphViewer node map
There are still some todo's and plenty of idea's for future work, but I decided to merge the current PoC with 9b65520 |
As discussed in #156 the idea is currently to trigger this feature with a job option like
|
Follow up todo's and tasks (WIP)
|
made an initial demo notebook at unfortunately the job metadata visualization (including process graph viz) does not work in that online notebook viewer, so the explanation might not always make sense |
Follow up on #95 and #53. Develop a cross-back end process graph splitting implementation that is useful in the current aggregator setup. It should be based on the algorithm developed during the feasibility study.
The text was updated successfully, but these errors were encountered: