Skip to content

Commit 019cdfa

Browse files
authored
Merge pull request #4748 from LAVEEN/gscresv
Removing support for maintenance_interval for reservations created by TAMs
2 parents 2d5b02d + 6fd2c17 commit 019cdfa

4 files changed

Lines changed: 2 additions & 19 deletions

File tree

examples/machine-learning/a3-highgpu-8g/README.md

Lines changed: 2 additions & 12 deletions
Original file line numberDiff line numberDiff line change
@@ -131,18 +131,12 @@ Cloud representative. Set them at approximately lines 33 and 34 of
131131
> skip this step and proceed to [manual reservation creation](#manual-creation-of-reservation).
132132

133133
Set the deployment variable `a3_reservation_name` at approximately line 38 of
134-
`ml-slurm-a3-2-cluster.yaml` to the reservation name provided by Google. The
135-
value for `a3_maintenance_interval` should also be set as directed by Google
136-
staff. A common setting is `PERIODIC`, shown below, but this value must be
137-
confirmed with Google staff.
134+
`ml-slurm-a3-2-cluster.yaml` to the reservation name provided by Google.
138135

139136
```yaml
140137
# a3_reservation_name must be specified; if Google staff have provided you
141138
# with a reservation name, use it. Otherwise supply user-created reservation.
142139
a3_reservation_name: reservation-name-provided-by-google
143-
# a3_maintenance_interval should be empty string by default; if Google staff
144-
# have created a reservation, they will also provide a3_maintenance_interval
145-
a3_maintenance_interval: PERIODIC
146140
```
147141

148142
### Manual creation of reservation
@@ -167,18 +161,14 @@ gcloud compute reservations create a3-reservation-0 \
167161

168162
This reservation be must be specified when creating VMs with matching parameters
169163
(e.g. a3-highgpu-8g VM in configured zone). If you executed the command above
170-
without modification, you may leave `a3_reservation_name` and
171-
`a3_maintenance_interval` at their default values in
164+
without modification, you may leave `a3_reservation_name` at their default values in
172165
`ml-slurm-a3-2-cluster.yaml`. Otherwise, ensure that the reservation name in the
173166
blueprint matches the name of the user-created reservation.
174167

175168
```yaml
176169
# a3_reservation_name must be specified; if Google staff have provided you
177170
# with a reservation name, use it. Otherwise supply user-created reservation.
178171
a3_reservation_name: a3-reservation-0
179-
# a3_maintenance_interval should be empty string by default; if Google staff
180-
# have created a reservation, they will also provide a3_maintenance_interval
181-
a3_maintenance_interval: ""
182172
```
183173

184174
### Using Spot VM or DWS Flex

examples/machine-learning/a3-highgpu-8g/ml-slurm-a3-2-cluster.yaml

Lines changed: 0 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -45,9 +45,6 @@ vars:
4545
a3_dws_flex_enabled: false
4646
a3_enable_spot_vm: false
4747

48-
# a3_maintenance_interval should be empty string by default; if Google staff
49-
# have created a reservation, they will also provide a3_maintenance_interval
50-
a3_maintenance_interval: ""
5148
# network parameters must match base blueprint deployment_name!
5249
# these values are accurate if deployment_name was not modified from example
5350
network_name_system: slurm-a3-base-sysnet
@@ -188,7 +185,6 @@ deployment_groups:
188185
enable_spot_vm: $(vars.a3_enable_spot_vm)
189186
dws_flex:
190187
enabled: $(vars.a3_dws_flex_enabled)
191-
maintenance_interval: $(vars.a3_maintenance_interval)
192188
node_count_static: $(vars.a3_static_cluster_size)
193189
node_count_dynamic_max: 0
194190
disk_type: pd-ssd

examples/machine-learning/a3-megagpu-8g/a3mega-slurm-blueprint.yaml

Lines changed: 0 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -27,7 +27,6 @@ vars:
2727
source_image_family: ubuntu-accelerator-2204-amd64-with-nvidia-570
2828
source_image_project_id: [ubuntu-os-accelerator-images]
2929
a3mega_partition_name: a3mega
30-
a3mega_maintenance_interval: ""
3130
local_mount_homefs: /home
3231
instance_image:
3332
family: $(vars.final_image_family)
@@ -549,7 +548,6 @@ deployment_groups:
549548
enable_spot_vm: $(vars.a3mega_enable_spot_vm)
550549
dws_flex:
551550
enabled: $(vars.a3mega_dws_flex_enabled)
552-
maintenance_interval: $(vars.a3mega_maintenance_interval)
553551
startup_script: $(a3mega_startup.startup_script)
554552

555553
- id: a3mega_partition

examples/machine-learning/a3-megagpu-8g/a3mega-slurm-deployment.yaml

Lines changed: 0 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -31,7 +31,6 @@ vars:
3131
disk_size_gb: 200
3232
final_image_family: slurm-a3mega
3333
slurm_cluster_name: a3mega
34-
a3mega_maintenance_interval: ""
3534
a3mega_cluster_size: 2 # supply cluster size
3635
a3mega_reservation_name: "" # supply reservation name
3736
# Additional provisioning models (pick only one), can be used to substitute `a3mega_reservation_name`:

0 commit comments

Comments
 (0)