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

Reducing resource requests #42

Open
wants to merge 3 commits into
base: main
Choose a base branch
from
Open

Reducing resource requests #42

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
c8c2023
Took a first pass at reducing the resources to be requested for the v…
drewoldag Sep 10, 2024
38a11ad
Need to update the `n_workers` parameter in the reprojection config a…
drewoldag Sep 10, 2024
105afdb
Additional tweaks to the klone config.
drewoldag Sep 11, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion example_runtime_config.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,7 +30,7 @@ overwrite = false

[apps.reproject_wu]
# Number of processors to use for parallelizing the reprojection
n_workers = 32
n_workers = 8

# The name of the observation site to use for reflex correction
observation_site = "ctio"
Expand Down
24 changes: 13 additions & 11 deletions src/kbmod_wf/resource_configs/klone_configuration.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,8 +8,8 @@
walltimes = {
"compute_bigmem": "01:00:00",
"large_mem": "04:00:00",
"sharded_reproject": "04:00:00",
"gpu_max": "08:00:00",
"sharded_reproject": "01:00:00",
"gpu_max": "01:00:00",
Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I reduced the time requested for each of these. @DinoBektesevic I think that 1hr should generally be enough to finish a search, but let me know if this should be pushed back up.

}


Expand All @@ -21,7 +21,7 @@ def klone_resource_config():
os.path.join("/gscratch/dirac/kbmod/workflow/run_logs", datetime.date.today().isoformat())
),
run_dir=os.path.join("/gscratch/dirac/kbmod/workflow/run_logs", datetime.date.today().isoformat()),
retries=1,
retries=100,
Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Until we have a good way to catch and ignore pre-emption "failures" that would increment the retry counter, we can naively set the max retry number of something large.

executors=[
HighThroughputExecutor(
label="small_cpu",
Expand All @@ -30,19 +30,20 @@ def klone_resource_config():
partition="ckpt-g2",
account="astro",
min_blocks=0,
max_blocks=4,
max_blocks=16,
init_blocks=0,
parallelism=1,
nodes_per_block=1,
cores_per_node=1, # perhaps should be 8???
mem_per_node=256, # In GB
mem_per_node=32, # In GB
Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This executor is only used by the pre-TNO workflow to convert the URI file into an ImageCollection. So we probably never needed anywhere near the memory that was requested.

exclusive=False,
walltime=walltimes["compute_bigmem"],
# Command to run before starting worker - i.e. conda activate <special_env>
worker_init="",
),
),
HighThroughputExecutor(
# This executor was used for the pre-TNO reprojection task
label="large_mem",
max_workers=1,
provider=SlurmProvider(
Expand All @@ -62,18 +63,19 @@ def klone_resource_config():
),
),
HighThroughputExecutor(
# This executor is used for reprojecting sharded WorkUnits
label="sharded_reproject",
max_workers=1,
provider=SlurmProvider(
partition="ckpt-g2",
account="astro",
min_blocks=0,
max_blocks=2,
max_blocks=16,
init_blocks=0,
parallelism=1,
nodes_per_block=1,
cores_per_node=32,
mem_per_node=128, # ~2-4 GB per core
cores_per_node=8,
mem_per_node=32, # ~2-4 GB per core
Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In this executor we're cranking up the maximum number of concurrent jobs running and decreasing the cores per node and memory.

exclusive=False,
walltime=walltimes["sharded_reproject"],
# Command to run before starting worker - i.e. conda activate <special_env>
Expand All @@ -87,12 +89,12 @@ def klone_resource_config():
partition="ckpt-g2",
account="escience",
min_blocks=0,
max_blocks=2,
max_blocks=10,
init_blocks=0,
parallelism=1,
nodes_per_block=1,
cores_per_node=2, # perhaps should be 8???
mem_per_node=512, # In GB
cores_per_node=1,
mem_per_node=64, # In GB
exclusive=False,
walltime=walltimes["gpu_max"],
# Command to run before starting worker - i.e. conda activate <special_env>
Expand Down
Loading