phoenix/litellm-mirror
0
0
Fork 1
mirror of https://github.com/BerriAI/litellm.git synced 2025-04-26 11:14:04 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

chore: increase livenessProbe and readinessProbe timeouts

Browse source 
- our litellm deployment keeps restarting because the pods fail readiness and liveness probes
- the default in k8s is 1s, the pod isn't always able to respond it time
- it seems to happen particularly often when the process flushes spend logs to the database
This commit is contained in:
Ashwin Madavan 2025-04-14 20:58:48 -04:00 • committed by GitHub
parent db857c74d4
commit 522da2be8a
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 2 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

2
deploy/charts/litellm-helm/templates/deployment.yaml
View file

@ -120,10 +120,12 @@ spec:
httpGet: httpGet:
path: /health/liveliness path: /health/liveliness
port: http port: http
timeoutSeconds: 10
readinessProbe: readinessProbe:
httpGet: httpGet:
path: /health/readiness path: /health/readiness
port: http port: http
timeoutSeconds: 10
# Give the container time to start up. Up to 5 minutes (10 * 30 seconds) # Give the container time to start up. Up to 5 minutes (10 * 30 seconds)
startupProbe: startupProbe:
httpGet: httpGet: