Fly.io deployment configuration and scaling

<p>Deploying a FastAPI application to Fly.io. Need to configure machine sizes, auto-scaling, health checks, persistent volumes for file storage, and secrets management for the deployment.</p>

<p>Configure <code>fly.toml</code> for a FastAPI deployment with auto-scaling:</p> <div class="highlight"><pre><span></span><code><span class="c1"># fly.toml</span> <span class="n">app</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="s1">'my-fastapi-app'</span> <span class="n">primary_region</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="s1">'ord'</span><span class="w"> </span><span class="c1"># Chicago — pick closest to users</span> <span class="k">[build]</span> <span class="w"> </span><span class="n">dockerfile</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="s1">'Dockerfile'</span> <span class="k">[env]</span> <span class="w"> </span><span class="n">APP_ENV</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="s1">'production'</span> <span class="w"> </span><span class="n">PORT</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="s1">'8000'</span> <span class="w"> </span><span class="c1"># Non-secret config here</span> <span class="k">[http_service]</span> <span class="w"> </span><span class="n">internal_port</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="mi">8000</span> <span class="w"> </span><span class="n">force_https</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="kc">true</span> <span class="w"> </span><span class="n">auto_stop_machines</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="kc">true</span><span class="w"> </span><span class="c1"># Stop when no traffic</span> <span class="w"> </span><span class="n">auto_start_machines</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="kc">true</span><span class="w"> </span><span class="c1"># Start on request</span> <span class="w"> </span><span class="n">min_machines_running</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="mi">0</span><span class="w"> </span><span class="c1"># Can scale to zero (saves money)</span> <span class="w"> </span><span class="n">processes</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="p">[</span><span class="s1">'app'</span><span class="p">]</span> <span class="w"> </span><span class="k">[http_service.concurrency]</span> <span class="w"> </span><span class="n">type</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="s1">'requests'</span> <span class="w"> </span><span class="n">hard_limit</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="mi">100</span><span class="w"> </span><span class="c1"># Max concurrent requests per machine</span> <span class="w"> </span><span class="n">soft_limit</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="mi">80</span><span class="w"> </span><span class="c1"># Start new machine when this is hit</span> <span class="k">[[http_service.checks]]</span> <span class="w"> </span><span class="n">grace_period</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="s1">'10s'</span> <span class="w"> </span><span class="n">interval</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="s1">'15s'</span> <span class="w"> </span><span class="n">method</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="s1">'GET'</span> <span class="w"> </span><span class="n">path</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="s1">'/health'</span> <span class="w"> </span><span class="n">timeout</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="s1">'5s'</span> <span class="k">[mounts]</span> <span class="w"> </span><span class="c1"># Persistent storage for file uploads</span> <span class="w"> </span><span class="n">source</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="s1">'uploads'</span> <span class="w"> </span><span class="n">destination</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="s1">'/app/uploads'</span> <span class="k">[[vm]]</span> <span class="w"> </span><span class="n">size</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="s1">'shared-cpu-1x'</span><span class="w"> </span><span class="c1"># 256MB RAM — good for APIs</span> <span class="w"> </span><span class="n">memory</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="s1">'512mb'</span> <span class="w"> </span><span class="n">cpu_kind</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="s1">'shared'</span> <span class="w"> </span><span class="n">cpus</span><span class="w"> </span><span class="o">=</span><span class="w"> </span><span class="mi">1</span> </code></pre></div> <div class="highlight"><pre><span></span><code><span class="c1"># Deploy</span> fly<span class="w"> </span>deploy <span class="c1"># Set secrets (encrypted at rest, injected as env vars)</span> fly<span class="w"> </span>secrets<span class="w"> </span><span class="nb">set</span><span class="w"> </span><span class="se">\</span> <span class="w"> </span><span class="nv">DATABASE_URL</span><span class="o">=</span><span class="s2">"postgresql+asyncpg://user:pass@host/db"</span><span class="w"> </span><span class="se">\</span> <span class="w"> </span><span class="nv">OPENAI_API_KEY</span><span class="o">=</span><span class="s2">"sk-..."</span><span class="w"> </span><span class="se">\</span> <span class="w"> </span><span class="nv">REDIS_URL</span><span class="o">=</span><span class="s2">"redis://..."</span> <span class="c1"># Scale machines manually</span> fly<span class="w"> </span>scale<span class="w"> </span>count<span class="w"> </span><span class="m">2</span><span class="w"> </span>--region<span class="w"> </span>ord fly<span class="w"> </span>scale<span class="w"> </span>vm<span class="w"> </span>performance-2x <span class="c1"># View logs</span> fly<span class="w"> </span>logs<span class="w"> </span>--app<span class="w"> </span>my-fastapi-app <span class="c1"># Open postgres console</span> fly<span class="w"> </span>postgres<span class="w"> </span>connect<span class="w"> </span>-a<span class="w"> </span>my-postgres-app <span class="c1"># Create persistent volume</span> fly<span class="w"> </span>volumes<span class="w"> </span>create<span class="w"> </span>uploads<span class="w"> </span>--region<span class="w"> </span>ord<span class="w"> </span>--size<span class="w"> </span><span class="m">10</span><span class="w"> </span><span class="c1"># 10GB</span> </code></pre></div> <p><code>auto_stop_machines = true</code> with <code>min_machines_running = 0</code> enables scale-to-zero (free tier friendly). <code>soft_limit</code> triggers scale-out before hard limit is hit. Fly uses Anycast routing — your machines are globally distributed automatically.</p>