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.

Sign up for GitHub

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

Jump to bottom

LB hangs with multiple live collections #1055

Closed
PhilMiller opened this issue Sep 15, 2020 · 2 comments
Closed

LB hangs with multiple live collections #1055

PhilMiller opened this issue Sep 15, 2020 · 2 comments
Assignees
@PhilMiller
Labels
type: bug

Comments

@PhilMiller
Copy link
Member

To Reproduce

$ mpiexec -n 2 "./test_model_per_collection_extended" --vt_lb_name GreedyLB --vt_lb and watch hang detection fire

Describe the bug

Two collections synchronize, only one gets released post-LB - i.e. in --vt_debug_lb output, num_calls == 2 while num_release == 1 after LB finishes.

Expected behavior

Releases wouldn't be counted, since LB should only run once.

Platform (please complete the following information):

  • Commit ID:
commit 152c8dca1f19e80a53e2df3033cf57b2fdcf5a75 (HEAD -> develop, origin/develop, origin/HEAD)
Merge: 3557f93b 95b67eed
Author: Jonathan Lifflander <jliffla@sandia.gov>
Date:   Sat Sep 12 20:51:41 2020 -0700

    Merge pull request #1038 from DARMA-tasking/1033-subphase-getter
    
    1033 Add getter for subphase on element stats

Additional context
Discovered while developing #965
@PhilMiller PhilMiller self-assigned this Sep 15, 2020
PhilMiller added a commit that referenced this issue Sep 15, 2020
@PhilMiller
#1055: Make post-LB release of execution uncounted, since invocation …
daceac9 
…is singular
@PhilMiller PhilMiller mentioned this issue Sep 15, 2020
Closed
PhilMiller added a commit that referenced this issue Sep 16, 2020
@PhilMiller
#1055: Preserve ability to look up object -> proxy mapping outside LB…
0671841 
… (i.e. during model testing)
PhilMiller added a commit that referenced this issue Sep 18, 2020
@PhilMiller
#1055: Make post-LB release of execution uncounted, since invocation …
d6aa84b 
…is singular
PhilMiller added a commit that referenced this issue Sep 18, 2020
@PhilMiller
#1055: Preserve ability to look up object -> proxy mapping outside LB…
8b5a7fd 
… (i.e. during model testing)
PhilMiller added a commit that referenced this issue Sep 21, 2020
@PhilMiller
#1055: Make post-LB release of execution uncounted, since invocation …
4c6b032 
…is singular
PhilMiller added a commit that referenced this issue Sep 21, 2020
@PhilMiller
#1055: Preserve ability to look up object -> proxy mapping outside LB…
26939aa 
… (i.e. during model testing)
PhilMiller added a commit that referenced this issue Sep 22, 2020
@PhilMiller
#1055: Make post-LB release of execution uncounted, since invocation …
660620a 
…is singular
PhilMiller added a commit that referenced this issue Sep 22, 2020
@PhilMiller
#1055: Preserve ability to look up object -> proxy mapping outside LB…
377f471 
… (i.e. during model testing)
PhilMiller added a commit that referenced this issue Sep 22, 2020
@PhilMiller
#1055: Avoid use of waitLBCollective
9738db9
@lifflander lifflander mentioned this issue Sep 22, 2020
Open
@PhilMiller PhilMiller mentioned this issue Sep 25, 2020
Closed
PhilMiller added a commit that referenced this issue Sep 25, 2020
@PhilMiller
#1055: Make post-LB release of execution uncounted, since invocation …
d64ab7d 
…is singular
PhilMiller added a commit that referenced this issue Sep 25, 2020
@PhilMiller
#1055: Preserve ability to look up object -> proxy mapping outside LB…
817030a 
… (i.e. during model testing)
PhilMiller added a commit that referenced this issue Sep 25, 2020
@PhilMiller
#1055: Avoid use of waitLBCollective
4eddbe7
PhilMiller added a commit that referenced this issue Sep 25, 2020
@PhilMiller
#1055: Delete waitLBCollective and synced_in_lb_
d0f9813
@PhilMiller PhilMiller mentioned this issue Sep 25, 2020
Merged
@PhilMiller
Copy link
Member Author

#1123 should make this moot.

@PhilMiller
Copy link
Member Author

mpiexec "-n" "2" "./collection_extended" "--gtest_filter=TestModelPerCollection.test_model_per_collection_1" --vt_lb --vt_lb_name GreedyLB

Now passes, so this can be closed.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@PhilMiller PhilMiller

Labels
type: bug
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

#1055 & #965: LB Refactoring, fix a hang with multiple collection DARMA-tasking/vt
1 participant
@PhilMiller