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

planner: simplify accounting during migration #427

Merged
merged 1 commit into from
Apr 18, 2024
Merged

Conversation

csegarragonz
Copy link
Collaborator

There was a bug in the accounting logic during a DIST_CHANGE request. At this point, we need to release the slots of the to-be migrated function, and claim them in the migrated-to hosts.

The current logic assumed that, given two decisions, they would always start being similar, and then diverge (or even diverge all the way through).

This is not the case. Take for example the following situation where A has 5 slots:

Old, New
B, A
B, A
B, A
C, A
C, A
A, C
A, C

Using our accounting logic, for a while we would have 7 slots assigned to A. This would break our assertions (did not catch it because it only happens in Release builds), but was breaking the MPI port assignation.

There was a bug in the accounting logic during a DIST_CHANGE request. At
this point, we need to release the slots of the to-be migrated function,
and claim them in the migrated-to hosts.

The current logic assumed that, given two decisions, they would always
start being similar, and then diverge (or even diverge all the way
through).

This is not the case. Take for example the following situation where A
has 5 slots:

```
Old, New
B, A
B, A
B, A
C, A
C, A
A, C
A, C
```

Using our accounting logic, for a while we would have 7 slots assigned
to A. This would break our assertions (did not catch it because it only
happens in Release builds), but was breaking the MPI port assignation.
Copy link

codecov bot commented Apr 18, 2024

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 82.18%. Comparing base (da5fee2) to head (0145420).

Additional details and impacted files
@@            Coverage Diff             @@
##             main     #427      +/-   ##
==========================================
- Coverage   82.25%   82.18%   -0.08%     
==========================================
  Files         115      115              
  Lines        7586     7584       -2     
==========================================
- Hits         6240     6233       -7     
- Misses       1346     1351       +5     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

@csegarragonz csegarragonz merged commit 952811a into main Apr 18, 2024
12 checks passed
@csegarragonz csegarragonz deleted the planner-fix branch April 18, 2024 09:57
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant