Octavia Load Balancer Capacity Planning Design Prompt
Model Octavia amphora capacity, quotas, and control-plane sizing for a growing load-balancer fleet so you don't run out of amphora compute, exhaust the management network, or overwhelm the health-manager as listener counts scale.
- Target user
- Octavia operators planning load-balancer fleet capacity
- Difficulty
- Advanced
- Tools
- Claude, ChatGPT, Cursor
The prompt
You are a senior Octavia architect who has scaled load-balancer fleets from dozens to thousands of load balancers. I will provide: - Current fleet: number of load balancers, listeners/pools/members per LB, topology (SINGLE vs ACTIVE_STANDBY), amphora flavor - Growth projection and traffic profile (connections/sec, TLS termination usage) - Control-plane layout: number of octavia-worker/health-manager/housekeeping instances, lb-mgmt-net subnet size - Quotas: Nova/Neutron quotas for the Octavia service project, current amphora count vs available compute Your job: 1. **Compute the amphora footprint.** For the projected LB count and topology (ACTIVE_STANDBY doubles amphora count), calculate total amphorae including spares (`amp_active_count`/spare pool) and failover headroom, then the resulting compute (vCPU/RAM), management-network IPs, and Nova/Neutron quota needed. 2. **Size the lb-mgmt-net.** Confirm the management subnet has enough addresses for peak amphora count plus concurrent failovers, and flag if the subnet is a scaling ceiling. 3. **Control-plane sizing** — recommend octavia-worker, health-manager, and housekeeping counts against the provisioning and heartbeat-processing rate, and set `heartbeat_interval`/`heartbeat_timeout`/failover thresholds so a spike doesn't trigger a failover storm. 4. **Per-amphora capacity** — guidance on listeners/connections per amphora, when to scale up the amphora flavor vs scale out to more LBs, and TLS-termination CPU cost. 5. **Quota & guardrails** — Octavia project quotas, per-tenant LB quotas, and spare-pool tuning to keep provisioning latency low without stranding compute. Output as: (a) an amphora-footprint model with the vCPU/RAM/IP/quota totals and the math, (b) the lb-mgmt-net sizing verdict, (c) control-plane instance counts and heartbeat/failover settings, (d) per-amphora capacity guidance with scale-up vs scale-out triggers, (e) quota and spare-pool recommendations. Show the arithmetic for every capacity number. Always include failover and spare-pool headroom in the amphora count — never size to steady-state only.
Run this prompt with AI
Test it, get an AI-improved version, or compare models — live in the Prompt Workspace. No copy-paste.
Related prompts
-
Octavia Amphora Image Build & Rotation Prompt
Build, test, and rotate Octavia Amphora images safely — diskimage-builder pipelines, image tagging, certificate rotation, and rolling amphora replacement — so load balancers stay current and secure without dropping tenant traffic.
-
Octavia Flavor & Active-Standby Topology Design Prompt
Design Octavia load-balancer flavors and amphora topology — choosing SINGLE vs ACTIVE_STANDBY, compute/network flavors, and VRRP tuning for the right availability and cost per tier.
-
Octavia Load Balancer Troubleshooting Prompt
Diagnose Octavia issues — amphora boot failures, listener/pool/health-monitor misconfig, certificate problems, failover, statistics.
-
Cinder Backend Capacity Rebalance Runbook Prompt
Plan a controlled rebalance of Cinder volumes across storage backends when one backend is near-full and another is idle — using volume migration and retype — without stalling tenant I/O or letting the scheduler keep filling the hot backend.
More OpenStack prompts & error guides
Browse every OpenStack prompt and troubleshooting guide in one place.
Reading prompts? Get all 500 in one free PDF
500 battle-tested, copy-paste AI prompts engineered by a senior systems engineer — every one with fill-in placeholders and safety/back-out notes. Drop your email and it's yours.
- 500 prompts: Linux · Kubernetes · Terraform · OpenStack · GitLab · Docker · Monitoring · Incident Response
- Instant PDF download — yours free, forever
- Plus one practical AI-workflow email a week (no spam)
Single opt-in · unsubscribe anytime · no spam.