Closed YongzuoLi-NOAA closed 1 year ago
@travissluka @guillaumevernieres Did anyone use parameters_bump_cor_nicas_scales in testinput/3dvarbump.yml for soca_var.x successfully ? @ShastriPaturi
Those tests are not designed to work together. Look at the geometry resolution in the yaml files and you'll see that the 3dvarbump test runs on a different lower resolution grid
Hi Travis,
Thank you for your response.
Yongzuo
On Wed, Sep 20, 2023 at 12:21 AM Travis Sluka @.***> wrote:
Closed #2 https://github.com/JCSDA/soca/issues/2 as not planned.
— Reply to this email directly, view it on GitHub https://github.com/JCSDA/soca/issues/2#event-10417639361, or unsubscribe https://github.com/notifications/unsubscribe-auth/AT2VRX5ZSCLOJLENASKDSTLX3JVOFANCNFSM6AAAAAA4ZBTJTU . You are receiving this because you authored the thread.Message ID: @.***>
Description
(1) I manually repeated the ctest for 3dvarbump successfully. cd /work2/noaa/da/yongzuo/testv0/test_workdir/3dvarbump "/apps/gcc-10.2.0/openmpi-4.0.4/openmpi-4.0.4/bin/mpiexec" "-n" "2" "/work2/noaa/marine/yli/JEDI/soca-20230731/soca_build/bin/soca_var.x" "testinput/3dvarbump.yml"
(2) I replaced parameters_bump_cor_nicas with parameters_bump_cor_nicas_scales in testinput/3dvarbump.yml Ran soca_var.x again as (1) and got the error below (crashed much earlier than my HAFS-MOM6 case). I will wait for you to pass this ctest to make sure the software is correct, right ?
Info : --- Read local NICAS parameters Info : Read NICAS local data Info : Read NICAS data of task 1 from task 1 Info : Block: socn Info : Component: 1 Info : Block: tocn Info : Component: 1 Info : Block: uocn Info : Component: 1 Info : Block: vocn Info : Component: 1 !!! ABORT in nicas_read_local on task #000002: wrong size for dimension nc0a
MPI_ABORT was invoked on rank 1 in communicator MPI_COMM_WORLD with errorcode 1.