Skip to content

Fix provisioning 1x1 and 1x1x1 topologies #595

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

Merged
merged 2 commits into from
Aug 19, 2025
Merged

Fix provisioning 1x1 and 1x1x1 topologies #595

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
42cb07b
fix: provisioning 1t tpu topologies
scaliby Aug 18, 2025
91c9127
style: reformat
scaliby Aug 18, 2025
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
15 changes: 11 additions & 4 deletions src/xpk/core/nodepool.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -32,6 +32,8 @@
create_or_update_cluster_configmap,
)
from .system_characteristics import AcceleratorType
from functools import reduce
from operator import mul

CLOUD_PLATFORM_AUTH_SCOPE_URL = (
'"https://www.googleapis.com/auth/cloud-platform"'
Expand Down Expand Up @@ -279,16 +281,21 @@ def run_gke_node_pool_create_command(
)
if system.accelerator_type == AcceleratorType['TPU']:
command += f' --node-version={gke_node_pool_version}'
topology_product = reduce(
mul, (int(x) for x in system.topology.split('x')), 1
)
if capacity_type == CapacityType.FLEX_START:
command += ' --num-nodes=0'
else:
elif topology_product > 1:
command += f' --num-nodes={system.vms_per_slice}'
command += ' --placement-type=COMPACT --max-pods-per-node 15'
command += (
f' --scopes=storage-full,gke-default,{CLOUD_PLATFORM_AUTH_SCOPE_URL}'
)
command += f' --tpu-topology={system.topology}'
command += f' {args.custom_tpu_nodepool_arguments}'

if topology_product > 1:
command += ' --placement-type=COMPACT --max-pods-per-node 15'
command += f' --tpu-topology={system.topology}'
command += f' {args.custom_tpu_nodepool_arguments}'
elif system.accelerator_type == AcceleratorType['GPU']:
subnet_prefix = f'{args.cluster}-{zone_to_region(args.zone)}'
if capacity_type == CapacityType.FLEX_START:
Expand Down
Loading