Skip to content

IVS-944 Memory budget limits #327

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

Open
Ghesselink wants to merge 1 commit into
base: development
Choose a base branch
from
+35 −7
Open

IVS-944 Memory budget limits #327

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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 Makefile
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -31,7 +31,7 @@ stop:
REGISTRY ?= localhost:5000
WORKERS ?= 2
ENV_FILE ?= .env
SWARM_VARS = REGISTRY CERTBOT_DOMAIN CERTBOT_EMAIL NFS_SERVER_IP NFS_BASE_PATH WORKER_CPU_LIMIT WORKER_MEMORY_LIMIT WORKER_CPU_RESERVATION WORKER_MEMORY_RESERVATION
SWARM_VARS = REGISTRY CERTBOT_DOMAIN CERTBOT_EMAIL NFS_SERVER_IP NFS_BASE_PATH WORKER_CPU_LIMIT WORKER_MEMORY_LIMIT WORKER_CPU_RESERVATION WORKER_MEMORY_RESERVATION WORKER_REPLICAS BACKEND_MEMORY_LIMIT SCHEDULER_MEMORY_LIMIT AV_WORKER_MEMORY_LIMIT
SWARM_ENV = ENV_FILE="$(ENV_FILE)" $(foreach v,$(SWARM_VARS),$(v)="$(shell grep '^$(v)=' $(ENV_FILE) | head -1 | cut -d= -f2-)")

start-swarm:
Expand Down
31 changes: 29 additions & 2 deletions docker-compose.swarm.nodb.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,6 +23,9 @@ services:
- validate
deploy:
replicas: 1
resources:
limits:
memory: 256M
placement:
constraints: [node.role == manager]
restart_policy:
Expand All @@ -43,6 +46,9 @@ services:
- validate
deploy:
replicas: 1
resources:
limits:
memory: ${BACKEND_MEMORY_LIMIT}
placement:
constraints: [node.role == manager]
restart_policy:
Expand All @@ -69,7 +75,7 @@ services:
networks:
- validate
deploy:
replicas: 2
replicas: ${WORKER_REPLICAS}
# No placement constraint - workers run on any node
resources:
limits:
Expand Down Expand Up @@ -97,7 +103,7 @@ services:
replicas: 1
resources:
limits:
memory: 4G
memory: ${AV_WORKER_MEMORY_LIMIT}
# No placement constraint - workers run on any node
restart_policy:
condition: on-failure
Expand All @@ -118,6 +124,9 @@ services:
- validate
deploy:
replicas: 1
resources:
limits:
memory: ${SCHEDULER_MEMORY_LIMIT}
placement:
constraints: [node.role == manager]
restart_policy:
Expand All @@ -135,6 +144,9 @@ services:
- validate
deploy:
replicas: 1
resources:
limits:
memory: 512M
placement:
constraints: [node.role == manager]
restart_policy:
Expand Down Expand Up @@ -163,6 +175,9 @@ services:
- validate
deploy:
replicas: 1
resources:
limits:
memory: 512M
placement:
constraints: [node.role == manager]
restart_policy:
Expand All @@ -187,6 +202,9 @@ services:
- validate
deploy:
replicas: 1
resources:
limits:
memory: 1G
placement:
constraints: [node.role == manager]
restart_policy:
Expand Down Expand Up @@ -215,6 +233,9 @@ services:
- validate
deploy:
replicas: 1
resources:
limits:
memory: 512M
placement:
constraints: [node.role == manager]
restart_policy:
Expand All @@ -231,6 +252,9 @@ services:
- validate
deploy:
mode: global # one per node; Prometheus discovers via tasks.node_exporter
resources:
limits:
memory: 128M
restart_policy:
condition: on-failure
delay: 5s
Expand All @@ -245,6 +269,9 @@ services:
- validate
deploy:
replicas: 1
resources:
limits:
memory: 512M
restart_policy:
condition: on-failure
delay: 5s
Expand Down
9 changes: 5 additions & 4 deletions docker/backend/worker-beat-entrypoint.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,7 +18,8 @@ do
done
echo "DB is ready."

CELERY_CONCURRENCY=${CELERY_CONCURRENCY:-6} # default 6 worker processes
echo "Celery concurrency: $CELERY_CONCURRENCY"

celery --app=core worker -Q celery --loglevel=info --concurrency $CELERY_CONCURRENCY --task-events --hostname=worker@%n --beat --scheduler django_celery_beat.schedulers:DatabaseScheduler
# Beat-only: this service schedules periodic tasks and no longer consumes the
# celery queue itself. Task capacity lives in the worker service, where it is
# budgeted (resources.limits x CELERY_CONCURRENCY); the previous embedded worker
# added 4 unbudgeted task slots and peaked at 8 GB RSS without any memory limit.
celery --app=core beat --loglevel=info --scheduler django_celery_beat.schedulers:DatabaseScheduler
Loading