fix: eliminate memory overcommit to prevent node OOM crashes
Set requests = limits (Guaranteed QoS) across LimitRange defaults and explicit pod resources. Node2 crashed 2026-03-14 from 250% memory overcommit (61GB limits on 24GB node). Changes: - LimitRange: default = defaultRequest for all 6 tiers - Grafana: 3 → 2 replicas - Grampsweb: document why replicas=0 - Prometheus: 1Gi/4Gi → 3Gi/3Gi - OpenClaw: 512Mi/2Gi → 768Mi/768Mi - Immich server: 256Mi/2Gi → 512Mi/512Mi - Immich postgresql: 256Mi/1Gi → 512Mi/512Mi - Calibre: 256Mi/1536Mi → 256Mi/256Mi - Linkwarden: 256Mi/1536Mi → 768Mi/768Mi - N8N: 256Mi/1Gi → 512Mi/512Mi - MySQL cluster: 1Gi/3-4Gi → 2Gi/2Gi - pg-cluster (CNPG): 512Mi/4Gi → 512Mi/512Mi - DBaaS ResourceQuota limits.memory: 64Gi → 12Gi [ci skip]
This commit is contained in:
parent
27fa8ea18f
commit
2be858f616
10 changed files with 70 additions and 30 deletions
|
|
@ -38,6 +38,10 @@ variable "forgejo_api_token" {
|
|||
type = string
|
||||
sensitive = true
|
||||
}
|
||||
variable "claude_memory_api_key" {
|
||||
type = string
|
||||
sensitive = true
|
||||
}
|
||||
variable "nfs_server" { type = string }
|
||||
|
||||
|
||||
|
|
@ -172,6 +176,15 @@ resource "kubernetes_config_map" "openclaw_config" {
|
|||
}
|
||||
}
|
||||
}
|
||||
plugins = {
|
||||
allow = ["memory-api"]
|
||||
slots = {
|
||||
memory = "memory-api"
|
||||
}
|
||||
load = {
|
||||
paths = ["/home/node/.openclaw/extensions"]
|
||||
}
|
||||
}
|
||||
commands = {
|
||||
native = true
|
||||
nativeSkills = true
|
||||
|
|
@ -442,6 +455,17 @@ resource "kubernetes_deployment" "openclaw" {
|
|||
done
|
||||
fi
|
||||
|
||||
# Install memory-api plugin from GitHub (always pull latest)
|
||||
if [ -d /openclaw-home/extensions/memory-api/.git ]; then
|
||||
(cd /openclaw-home/extensions/memory-api && git pull --ff-only) || true
|
||||
else
|
||||
rm -rf /openclaw-home/extensions/memory-api
|
||||
git clone --depth 1 git@github.com:ViktorBarzin/claude-memory-mcp.git /tmp/claude-memory-mcp
|
||||
mkdir -p /openclaw-home/extensions/memory-api
|
||||
cp -r /tmp/claude-memory-mcp/openclaw-plugin/* /openclaw-home/extensions/memory-api/
|
||||
rm -rf /tmp/claude-memory-mcp
|
||||
fi
|
||||
|
||||
# Create required directories (owned by node user, UID 1000)
|
||||
mkdir -p /openclaw-home/agents/main/sessions /openclaw-home/credentials /openclaw-home/canvas /openclaw-home/devices /openclaw-home/cron /openclaw-home/cc-skills /openclaw-home/memory
|
||||
chown -R 1000:1000 /openclaw-home
|
||||
|
|
@ -570,6 +594,15 @@ resource "kubernetes_deployment" "openclaw" {
|
|||
name = "SLACK_WEBHOOK_URL"
|
||||
value = var.openclaw_skill_secrets["slack_webhook"]
|
||||
}
|
||||
# Memory API
|
||||
env {
|
||||
name = "MEMORY_API_URL"
|
||||
value = "http://claude-memory.claude-memory.svc.cluster.local"
|
||||
}
|
||||
env {
|
||||
name = "MEMORY_API_KEY"
|
||||
value = var.claude_memory_api_key
|
||||
}
|
||||
# Python packages path for skills
|
||||
env {
|
||||
name = "PYTHONPATH"
|
||||
|
|
@ -597,11 +630,11 @@ resource "kubernetes_deployment" "openclaw" {
|
|||
}
|
||||
resources {
|
||||
limits = {
|
||||
memory = "2Gi"
|
||||
memory = "768Mi"
|
||||
}
|
||||
requests = {
|
||||
cpu = "100m"
|
||||
memory = "512Mi"
|
||||
memory = "768Mi"
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue