-
Notifications
You must be signed in to change notification settings - Fork 21
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
Changes for finer granularity #13
Open
davegill
wants to merge
83
commits into
leave_docker_images
Choose a base branch
from
regression+feature
base: leave_docker_images
Could not load branches
Branch not found: {{ refName }}
Loading
Could not load tags
Nothing to show
Loading
Are you sure you want to change the base?
Some commits from the old base branch may be removed from the timeline,
and old review comments may become outdated.
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
1. We are back to chem tests 1, 2, 5. There was no timing performance advantage to split the tests apart. Howevr, using a BIGGER macine, with more cores for parallel compilation, was important. 2. Zapped NMM. 3. Split apart the 10, 13, 14, 15 tests. These all have serial components that run quite slowly. The may need to be further split. 4. To support this whole splitting thing, remove the hard-coded "19", and actually count how many tests are being conducted. 5. Introduce the FEATURE flag, in anticipation of usage with Kelly's feature (restart) tests. 6. With Varun, rewrote the output "part.sh" file. This is to allow the code to be less hard-coded in the groovy scripts. This required calling the last_oinly_once script. 7. The "last_only_once.csh" script now has an argument. This is the directory where the various SUCCESS* files are held. modified: build.csh
1. Zap a debugging "HI" 2. Indent the last_only_once.csh call 3. Delineate stanzas with additional line break. modified: build.csh
@vlakshmanan-scala
|
modified: build.csh
The purpose for this re-ordering is due to time constraints. The chemistry tests take more than 3x as long to build as the other ARW tests. For $ concerns, all tests were moved to a cheaper (smaller) AWS instance. This smaller machine type had fewer proessors, and therefore compiled the chem tests too slowly. The decision was to bump up the machine type, but only for chemistry job. The easiest way to handle the split machine assignment in the groovy scripting is to make the chemistry tests the first job. modified: build.csh
This is MPI only, now. The first case is "basic". modified: build.csh
We need to have either ( SERIAL && OPENMP ) || ( SERIAL && MPI ) to call the last_only_once script. modified: build.csh
To test this on the MMM Classroom machines, I need to have the feature test one of the early tests, since the scripts are broken apart. The actual order does not matter, as long as the first test is chemistry (which requires a larger machine on AWS). modified: build.csh
Only one proc is supposed to change the permissions in the OUTPUT directory. So put the chmod within an if test. Then, when the code is run on the MMM classroom, all is OK, and also when run on AWS. modified: build.csh
In the single_end.csh, make sure to remove the image for wrf_regtest. If this is laying around on the MMM classroom machines, we can't then do another test. modified: build.csh
modified: build.csh
The call to the feature_test.csh schript now has an argument for the RUNDIR, which is used to output the correct name of the SUCCESS message. modified: build.csh
Fumble-fingered the quote for the feature_testing.csh call. modified: build.csh
modified: build.csh
modified: build.csh On branch regression+feature Your branch is up to date with 'origin/regression+feature'. You are currently cherry-picking commit 4a9d582. Changes to be committed: modified: build.csh
modified: build.csh
modified: build.csh
modified: build.csh
modified: build.csh
This call has a single argument. That means - clean things up. modified: build.csh
modified: build.csh
1. It needs to only be handled when this is a feature test 2. The feature_testing.csh needs to be prefaced by the docker commands modified: build.csh
modified: README_add_feature_test.md
modified: README_add_feature_test.md
new file: README_x11.md
Pull in the da_builds.csh file from the SCRIPTS directory. This is used by test_000m.csh modified: Dockerfile modified: Dockerfile-second_part modified: Dockerfile-sed
modified: build.csh
Where are those SUCCESS files??? modified: build.csh
modified: build.csh
Need to append home/ubuntu/wrf-stuff/wrf-coop/OUTPUT to the directory location to where to find the SUCCESS files. modified: build.csh
fix test_000m.csh location of the famous "foo" file modified: build.csh
Orig New -------- 1 1 2 2 11 3 13 4 21 5 22 6 This allows a grouping of jobs at the head of the line that all can benefit from using the more expensive AWS machines (more processors - but apparently more memory is the big deal). modified: build.csh
modified: build.csh
17AD done in real*8 No real need for specific move tests - no one does this EVER. modified: build.csh
For use with jenkins bringing back all SUCCESS_RUN_WRF files for comparison locally. modified: build.csh
modified: build.csh
modified: build.csh
modified: build.csh
This tests the combo of all_wrfvar and ./configure wrfplus ./compile wrfplus and ./configure 4dvar ./compile all_wrfvar modified: build.csh
…csh (#14) Added to the build.csh script to include new feature tests for diff_opt=2, km_opt=1,2,3, and nest starts later. These coincide with [PR#3 in wrf_feature_testing](davegill/wrf_feature_testing#3).
Already have nested cases, want a stochastic case. modified: build.csh
modified: build.csh
Move the 03ST case to a non-OpenMP suite modified: build.csh
We want to see a failure message if DA does not compile. modified: build.csh
The unique build number will allow multiple regtests to be run at the same time modified: build.csh
modified: build.csh
modified: build.csh
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
the tests apart. Howevr, using a BIGGER macine, with more cores for parallel compilation,
was important.
slowly. The may need to be further split.
how many tests are being conducted.
hard-coded in the groovy scripts. This required calling the last_oinly_once script.
various SUCCESS* files are held.
modified: build.csh