chore(deps): v2.8 — refresh pinned components

Bump low-risk pins to current upstream, validated end-to-end on a live
AL2023 instance (Docker Compose v5.1.4 installs and runs; all three
bumped images pull and serve 200 on their readiness/metrics endpoints):

- Prometheus v3.1.0 -> v3.12.0
- node_exporter v1.9.0 -> v1.11.1 (all compose files)
- Pushgateway v1.11.2 -> v1.11.3
- Docker Compose plugin v2.29.7 -> v5.1.4 (AL2/AL2023 binary install;
  Ubuntu/RHEL continue to use the distro docker-compose-plugin package)

README Components table synced; dropped stale "v2" wording from the
Compose label and installer comments.

Held back deliberately:
- Grafana stays at 11.2.2 (11->13 is a two-major jump needing a
  dashboard/datasource compatibility pass).
- DCGM exporter stays at 4.2.0-4.1.0 (newer tags unpullable on Docker
  29.x, #47; configurability tracked in #50).

Author: nicolaven

CHANGELOG.md

@@ -1,5 +1,22 @@
 # Changelog
 
+## v2.8 — 2026-06-04
+
+Dependency refresh. Bumps the low-risk pinned components to current
+upstream releases. No functional changes. Grafana is intentionally held
+at 11.2.2 (the 11→13 jump needs a separate dashboard-compatibility pass)
+and DCGM exporter stays at 4.2.0-4.1.0 (Docker 29.x pull constraint, #47;
+configurability tracked in #50).
+
+### Changed
+- Prometheus `v3.1.0` → `v3.12.0`
+- node_exporter `v1.9.0` → `v1.11.1` (all compose files)
+- Pushgateway `v1.11.2` → `v1.11.3`
+- Docker Compose plugin `v2.29.7` → `v5.1.4` (AL2 / AL2023 binary install;
+  Ubuntu/RHEL continue to use the distro `docker-compose-plugin` package)
+- README Components table synced; dropped stale "v2" wording from the
+  Compose label and installer comments.
+
 ## v2.7 — 2026-06-03
 
 EFA fabric metrics and a searchable CloudWatch Logs dashboard. No breaking

README.md

@@ -303,13 +303,13 @@ ParallelCluster HeadNode / PCS Login Node
 | Component | Version |
 |-----------|---------|
 | Grafana | 11.2.2 |
-| Prometheus | v3.1.0 |
-| Pushgateway | v1.11.2 |
-| Node Exporter | v1.9.0 |
+| Prometheus | v3.12.0 |
+| Pushgateway | v1.11.3 |
+| Node Exporter | v1.11.1 |
 | NGINX | 1.27-alpine |
 | NVIDIA DCGM Exporter | 4.2.0-4.1.0-ubuntu22.04 |
 | prometheus-slurm-exporter | 1.8.0 (ParallelCluster only) |
-| Docker Compose v2 | 2.29.7 |
+| Docker Compose | 5.1.4 |
 | CloudWatch Exporter | v0.16.0 |
 
 All images pinned — `latest` is never used.

compose/compute.gpu.yml

@@ -14,7 +14,7 @@ services:
     volumes:
       - '/:/host:ro,rslave'
       - '/var/lib/prometheus/node-exporter:/textfile:ro'
-    image: quay.io/prometheus/node-exporter:v1.9.0
+    image: quay.io/prometheus/node-exporter:v1.11.1
     command:
       - '--path.rootfs=/host'
       - '--collector.infiniband'

compose/compute.yml

@@ -13,7 +13,7 @@ services:
     volumes:
       - '/:/host:ro,rslave'
       - '/var/lib/prometheus/node-exporter:/textfile:ro'
-    image: quay.io/prometheus/node-exporter:v1.9.0
+    image: quay.io/prometheus/node-exporter:v1.11.1
     command:
       - '--path.rootfs=/host'
       - '--collector.infiniband'

compose/head.yml

@@ -10,7 +10,7 @@ services:
     network_mode: host
     pid: host
     restart: unless-stopped
-    image: prom/pushgateway:v1.11.2
+    image: prom/pushgateway:v1.11.3
 
   prometheus:
     container_name: prometheus
@@ -26,7 +26,7 @@ services:
       - '__MONITORING_HOME__/prometheus/rules:/etc/prometheus/rules:ro'
       - 'prometheus-data:/prometheus'
       - '/run/prometheus-ec2-creds:/run/prometheus-ec2-creds:ro'
-    image: prom/prometheus:v3.1.0
+    image: prom/prometheus:v3.12.0
     command:
       - '--config.file=/etc/prometheus/prometheus.yml'
       - '--storage.tsdb.path=/prometheus'
@@ -71,7 +71,7 @@ services:
     volumes:
       - '/:/host:ro,rslave'
       - '/var/lib/prometheus/node-exporter:/textfile:ro'
-    image: quay.io/prometheus/node-exporter:v1.9.0
+    image: quay.io/prometheus/node-exporter:v1.11.1
     command:
       - '--path.rootfs=/host'
       - '--collector.infiniband'

installer/common.sh

@@ -121,7 +121,7 @@ install_efa_collector() {
 # ---------------------------------------------------------------------------
 verify_docker() {
     docker --version >/dev/null 2>&1 || die "docker not installed"
-    docker compose version >/dev/null 2>&1 || die "docker compose v2 plugin not installed"
+    docker compose version >/dev/null 2>&1 || die "docker compose plugin not installed"
     log "docker: $(docker --version)"
     log "compose: $(docker compose version --short)"
 }

installer/os/alinux2.sh

@@ -2,16 +2,16 @@
 # shellcheck disable=SC2154  # PLATFORM_USER from installer/platform/platform.sh
 # Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
 # SPDX-License-Identifier: MIT-0
-# Amazon Linux 2 package installation (docker + compose v2 plugin).
+# Amazon Linux 2 package installation (docker + compose plugin).
 set -euo pipefail
 
 log "Installing docker on Amazon Linux 2"
 # Amazon Linux 2 has docker in its extras repo.
 amazon-linux-extras install -y docker
 yum -y install jq bc tar gzip
 
-# Docker Compose v2 is NOT in AL2 repos. Install as a plugin binary.
-COMPOSE_VERSION="v2.29.7"
+# Docker Compose is NOT in AL2 repos. Install as a plugin binary.
+COMPOSE_VERSION="v5.1.4"
 install -d -m 0755 /usr/libexec/docker/cli-plugins
 curl -fsSL "https://github.com/docker/compose/releases/download/${COMPOSE_VERSION}/docker-compose-linux-$(uname -m)" \
     -o /usr/libexec/docker/cli-plugins/docker-compose

installer/os/alinux2023.sh

@@ -19,7 +19,7 @@ dnf -y install bc || log "WARN: bc not available, cost scripts may be degraded"
 
 # AL2023 does not ship docker-compose-plugin in its default repos yet.
 # Install upstream plugin binary.
-COMPOSE_VERSION="v2.29.7"
+COMPOSE_VERSION="v5.1.4"
 install -d -m 0755 /usr/libexec/docker/cli-plugins
 curl -fsSL "https://github.com/docker/compose/releases/download/${COMPOSE_VERSION}/docker-compose-linux-$(uname -m)" \
     -o /usr/libexec/docker/cli-plugins/docker-compose