AIX Monitoring
AIX Monitoring Bundle
Full stack: AIX_exporter + Loki Agent — complete AIX monitoring bundle. Prometheus metrics + Loki error logs in one package. Save 13% vs buying separately. Grafana dashboard included.
Complete AIX Observability — Metrics + Logs in One Bundle
The AIX Monitoring Bundle pairs AIX_exporter and Loki Agent for AIX into a single package — the two pillars of modern observability for IBM Power Systems, with a 15% bundle discount. Performance metrics from libperfstat, errpt streams in Loki, all visualized in your Grafana stack.
One purchase, one support contract
AIX_exporter and Loki Agent for AIX, packaged together. Single license, single invoice, 15% discount vs buying separately. Same BFF format, same SRC subsystem patterns, same config style — your sysadmins learn one stack instead of two.
# installp -aXcd /tmp/aix-exporter.1.0.0.2.bff aix-exporter.rte
SUCCESS: aix-exporter.rte 1.0.0.2
# installp -aXcd /tmp/loki-agent.1.0.0.1.bff loki-agent
SUCCESS: loki-agent 1.0.0.1
# startsrc -s aix-exporter
0513-059 The aix-exporter Subsystem has been started.
Subsystem PID is 6291602.
# startsrc -s loki-agent
0513-059 The loki-agent Subsystem has been started.
Subsystem PID is 6291718.
# Both agents now active — metrics on :9100, logs pushed to Loki
Metrics and logs, the two pillars
Performance metrics from AIX_exporter — CPU, memory, disks, FC, NFS, LPAR config — paired with errpt streams from Loki Agent. Both pillars of observability in your Grafana stack — finally for IBM AIX. AIX_exporter exposes 137 metrics on a Prometheus endpoint, Loki Agent pushes structured errpt entries to Loki — they live side by side in the same Grafana.
Spot what numbers can’t tell you
Open one Grafana dashboard. CPU spikes at 02:47 on lpar-prod-01. Drop to the Loki panel right below — the errpt entry that triggered it is right there, same timestamp. Root cause analysis goes from minutes of grepping across hosts to seconds of scrolling.
FCP_ERR4 (LINK ERROR)Unified install on every LPAR
Both agents installed via standard installp, registered as SRC subsystems, auto-start on boot. One Ansible playbook handles both. Same DevOps muscle memory, same config patterns, same observability stack — no duplicate processes to maintain across the fleet.
---
# Ansible playbook — bundle deploy across the fleet
- hosts: aix_fleet
tasks:
- name: Install AIX_exporter BFF
shell: installp -aXcd {{ pkg_dir }}/aix-exporter.bff aix-exporter.rte
args: { creates: /etc/aix-exporter/aix-exporter.conf }
- name: Install Loki Agent BFF
shell: installp -aXYgd {{ pkg_dir }}/loki-agent.bff all
args: { creates: /etc/loki/loki-agent.conf }
- name: Start both subsystems
shell: startsrc -s {{ item }}
loop: [ aix-exporter, loki-agent ]
What’s in the bundle
AIX_exporter
22 native collectors, 137 metrics, 600+ time series per LPAR. Single C binary on libperfstat, OpenMetrics-compatible, ships with a 60-panel Grafana dashboard.
Loki Agent for AIX
Real-time errpt forwarding to Grafana Loki with structured labels (hostname, machine_id, class, type, resource). LogQL queries, multi-channel alerting, optional TLS.
Workflow
Use Cases
Incident replay
Reconstruct any outage from both signal types: the metric trajectory and the errpt cascade, side by side, exact timeline.
Root cause in seconds
Performance degradation visible in metrics, hardware fault visible in errpt — both in one pane of glass with shared timestamps.
Compliance & audit
Historical record of system state and events combined — answer auditor questions about both performance and incidents from a single queryable store.
Capacity & reliability
Long-term trends across CPU/memory/disk plus error rates per LPAR — anticipate hardware refreshes before failures cascade.
Bundle Pricing
| LPARs | Separate | Bundle | Savings |
|---|---|---|---|
| 1 – 10 | €150 / LPAR | €130 / LPAR | 13% |
| 11 – 50 | €127 / LPAR | €110 / LPAR | 13% |
| 50+ | €113 / LPAR | €98 / LPAR | 13% |
What’s included
- AIX_exporter BFF package (full license, 22 collectors)
- Loki Agent for AIX BFF package (full license, standard build — TLS variant available on request)
- Two ready-to-import Grafana dashboard JSON files (60 panels for metrics, errpt panels for logs)
- Ansible playbook for fleet-wide deployment of both agents
- Documentation, installation guide, integration examples
- First year of maintenance — updates + email support for both products
Specs & Requirements
| Operating System | IBM AIX 7.1, 7.2, 7.3 |
| Architecture | IBM Power Systems (POWER7 and later) |
| Footprint | < 8 MB RSS combined (5 MB exporter + 3 MB agent), < 1% CPU each |
| Network | One TCP port for Prometheus scrape (9100), outbound HTTP/HTTPS to Loki |
| Backends | Prometheus 2.x or 3.x · Grafana Loki 2.x or 3.x |
| Grafana | 9.x, 10.x, 11.x, 12.x |
Deploy on AIX in 10 minutes
Both agents follow the same pattern: BFF package, SRC subsystem, auto-start on boot. Install on one LPAR to validate, then roll out to the fleet via the included Ansible playbook.
# 1. Install both BFF packages installp -aXcd /path/to/aix-exporter.1.0.0.2.bff aix-exporter.rte installp -aXYgd /path/to/loki-agent.1.0.0.1.bff all # 2. (Optional) Edit configurations vi /etc/aix-exporter/aix-exporter.conf vi /etc/loki/loki-agent.conf # 3. Start both subsystems startsrc -s aix-exporter startsrc -s loki-agent # 4. Verify both endpoints curl http://localhost:9100/metrics | head {service_name="errpt"} | json # in Grafana Explore
Add the AIX_exporter target to your prometheus.yml, point Loki Agent at your Loki instance via /etc/loki/loki-agent.conf, import the two Grafana dashboard JSONs. Done.
Licensing
Perpetual license for both products — pay once, use forever. Per-LPAR pricing with volume discounts, 13–15% savings over individual purchase. First year of maintenance (updates + support) included for both. TLS variant of Loki Agent available on request after purchase.
Frequently Asked Questions
Can I buy just one product later if I start with the bundle?
Yes. The bundle license is composed of two independent product licenses — you can keep using either one if you decide to drop the other at renewal. We don’t lock licenses together.
Do I need both Prometheus AND Loki?
Yes — Prometheus stores time-series metrics (from AIX_exporter), Loki stores log streams (from Loki Agent). They run alongside each other and both feed Grafana. If you already have a Prometheus stack, adding Loki is straightforward. If you have neither, the included documentation walks you through both.
What’s the resource overhead of running both agents?
Combined: under 8 MB RSS (5 MB exporter + 3 MB agent) and well under 2% CPU on a typical LPAR. Both agents are single-threaded C binaries that read kernel interfaces (libperfstat, errlog) and emit data — no fork bombs, no JVM, no container runtime.
Can I deploy with Ansible?
Yes — a unified Ansible playbook is included that installs and configures both agents in one pass, with pre-flight checks and post-install verification.
What about the trial?
Free trials are available for each product individually. Combine the two free trials to evaluate the full bundle on your infrastructure before purchase, no time limit.
