fix(infra): increase Tempo memory to 1Gi, RL runner CPU limits to 8000m
Tempo was OOMKilled every ~30 min at 512Mi, causing web-gateway OTLP export errors. Increased to 1Gi. RL runner had cpu_request_overwrite_max_allowed=4000m but hyperopt RL jobs need 6000m for parallel PSO trials. Increased to 8000m. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
@@ -34,10 +34,11 @@ runners:
|
||||
image = "rust:1.89-slim"
|
||||
privileged = false
|
||||
node_selector_overwrite_allowed = ".*"
|
||||
cpu_request_overwrite_max_allowed = "4000m"
|
||||
cpu_limit_overwrite_max_allowed = "4000m"
|
||||
memory_request_overwrite_max_allowed = "16Gi"
|
||||
memory_limit_overwrite_max_allowed = "16Gi"
|
||||
# Hyperopt RL jobs need 6000m CPU + 20Gi mem (PSO parallel trials on L4 8 vCPU)
|
||||
cpu_request_overwrite_max_allowed = "8000m"
|
||||
cpu_limit_overwrite_max_allowed = "8000m"
|
||||
memory_request_overwrite_max_allowed = "24Gi"
|
||||
memory_limit_overwrite_max_allowed = "24Gi"
|
||||
poll_timeout = 600
|
||||
# Allow CI jobs to override runtimeClassName via KUBERNETES_RUNTIME_CLASS_NAME
|
||||
runtime_class_name_overwrite_allowed = ".*"
|
||||
|
||||
@@ -77,10 +77,10 @@ spec:
|
||||
resources:
|
||||
requests:
|
||||
cpu: 100m
|
||||
memory: 128Mi
|
||||
memory: 256Mi
|
||||
limits:
|
||||
cpu: 500m
|
||||
memory: 512Mi
|
||||
memory: 1Gi
|
||||
readinessProbe:
|
||||
httpGet:
|
||||
path: /ready
|
||||
|
||||
Reference in New Issue
Block a user