461 lines
20 KiB
HTML
461 lines
20 KiB
HTML
<!DOCTYPE HTML>
|
|
<html lang="en" class="light sidebar-visible" dir="ltr">
|
|
<head>
|
|
<!-- Book generated using mdBook -->
|
|
<meta charset="UTF-8">
|
|
<title>0012: LLM Routing Tiers - VAPORA Platform Documentation</title>
|
|
|
|
|
|
<!-- Custom HTML head -->
|
|
|
|
<meta name="description" content="Comprehensive documentation for VAPORA, an intelligent development orchestration platform built entirely in Rust.">
|
|
<meta name="viewport" content="width=device-width, initial-scale=1">
|
|
<meta name="theme-color" content="#ffffff">
|
|
|
|
<link rel="icon" href="../favicon.svg">
|
|
<link rel="shortcut icon" href="../favicon.png">
|
|
<link rel="stylesheet" href="../css/variables.css">
|
|
<link rel="stylesheet" href="../css/general.css">
|
|
<link rel="stylesheet" href="../css/chrome.css">
|
|
<link rel="stylesheet" href="../css/print.css" media="print">
|
|
|
|
<!-- Fonts -->
|
|
<link rel="stylesheet" href="../FontAwesome/css/font-awesome.css">
|
|
<link rel="stylesheet" href="../fonts/fonts.css">
|
|
|
|
<!-- Highlight.js Stylesheets -->
|
|
<link rel="stylesheet" id="highlight-css" href="../highlight.css">
|
|
<link rel="stylesheet" id="tomorrow-night-css" href="../tomorrow-night.css">
|
|
<link rel="stylesheet" id="ayu-highlight-css" href="../ayu-highlight.css">
|
|
|
|
<!-- Custom theme stylesheets -->
|
|
|
|
|
|
<!-- Provide site root and default themes to javascript -->
|
|
<script>
|
|
const path_to_root = "../";
|
|
const default_light_theme = "light";
|
|
const default_dark_theme = "dark";
|
|
</script>
|
|
<!-- Start loading toc.js asap -->
|
|
<script src="../toc.js"></script>
|
|
</head>
|
|
<body>
|
|
<div id="mdbook-help-container">
|
|
<div id="mdbook-help-popup">
|
|
<h2 class="mdbook-help-title">Keyboard shortcuts</h2>
|
|
<div>
|
|
<p>Press <kbd>←</kbd> or <kbd>→</kbd> to navigate between chapters</p>
|
|
<p>Press <kbd>S</kbd> or <kbd>/</kbd> to search in the book</p>
|
|
<p>Press <kbd>?</kbd> to show this help</p>
|
|
<p>Press <kbd>Esc</kbd> to hide this help</p>
|
|
</div>
|
|
</div>
|
|
</div>
|
|
<div id="body-container">
|
|
<!-- Work around some values being stored in localStorage wrapped in quotes -->
|
|
<script>
|
|
try {
|
|
let theme = localStorage.getItem('mdbook-theme');
|
|
let sidebar = localStorage.getItem('mdbook-sidebar');
|
|
|
|
if (theme.startsWith('"') && theme.endsWith('"')) {
|
|
localStorage.setItem('mdbook-theme', theme.slice(1, theme.length - 1));
|
|
}
|
|
|
|
if (sidebar.startsWith('"') && sidebar.endsWith('"')) {
|
|
localStorage.setItem('mdbook-sidebar', sidebar.slice(1, sidebar.length - 1));
|
|
}
|
|
} catch (e) { }
|
|
</script>
|
|
|
|
<!-- Set the theme before any content is loaded, prevents flash -->
|
|
<script>
|
|
const default_theme = window.matchMedia("(prefers-color-scheme: dark)").matches ? default_dark_theme : default_light_theme;
|
|
let theme;
|
|
try { theme = localStorage.getItem('mdbook-theme'); } catch(e) { }
|
|
if (theme === null || theme === undefined) { theme = default_theme; }
|
|
const html = document.documentElement;
|
|
html.classList.remove('light')
|
|
html.classList.add(theme);
|
|
html.classList.add("js");
|
|
</script>
|
|
|
|
<input type="checkbox" id="sidebar-toggle-anchor" class="hidden">
|
|
|
|
<!-- Hide / unhide sidebar before it is displayed -->
|
|
<script>
|
|
let sidebar = null;
|
|
const sidebar_toggle = document.getElementById("sidebar-toggle-anchor");
|
|
if (document.body.clientWidth >= 1080) {
|
|
try { sidebar = localStorage.getItem('mdbook-sidebar'); } catch(e) { }
|
|
sidebar = sidebar || 'visible';
|
|
} else {
|
|
sidebar = 'hidden';
|
|
}
|
|
sidebar_toggle.checked = sidebar === 'visible';
|
|
html.classList.remove('sidebar-visible');
|
|
html.classList.add("sidebar-" + sidebar);
|
|
</script>
|
|
|
|
<nav id="sidebar" class="sidebar" aria-label="Table of contents">
|
|
<!-- populated by js -->
|
|
<mdbook-sidebar-scrollbox class="sidebar-scrollbox"></mdbook-sidebar-scrollbox>
|
|
<noscript>
|
|
<iframe class="sidebar-iframe-outer" src="../toc.html"></iframe>
|
|
</noscript>
|
|
<div id="sidebar-resize-handle" class="sidebar-resize-handle">
|
|
<div class="sidebar-resize-indicator"></div>
|
|
</div>
|
|
</nav>
|
|
|
|
<div id="page-wrapper" class="page-wrapper">
|
|
|
|
<div class="page">
|
|
<div id="menu-bar-hover-placeholder"></div>
|
|
<div id="menu-bar" class="menu-bar sticky">
|
|
<div class="left-buttons">
|
|
<label id="sidebar-toggle" class="icon-button" for="sidebar-toggle-anchor" title="Toggle Table of Contents" aria-label="Toggle Table of Contents" aria-controls="sidebar">
|
|
<i class="fa fa-bars"></i>
|
|
</label>
|
|
<button id="theme-toggle" class="icon-button" type="button" title="Change theme" aria-label="Change theme" aria-haspopup="true" aria-expanded="false" aria-controls="theme-list">
|
|
<i class="fa fa-paint-brush"></i>
|
|
</button>
|
|
<ul id="theme-list" class="theme-popup" aria-label="Themes" role="menu">
|
|
<li role="none"><button role="menuitem" class="theme" id="default_theme">Auto</button></li>
|
|
<li role="none"><button role="menuitem" class="theme" id="light">Light</button></li>
|
|
<li role="none"><button role="menuitem" class="theme" id="rust">Rust</button></li>
|
|
<li role="none"><button role="menuitem" class="theme" id="coal">Coal</button></li>
|
|
<li role="none"><button role="menuitem" class="theme" id="navy">Navy</button></li>
|
|
<li role="none"><button role="menuitem" class="theme" id="ayu">Ayu</button></li>
|
|
</ul>
|
|
<button id="search-toggle" class="icon-button" type="button" title="Search (`/`)" aria-label="Toggle Searchbar" aria-expanded="false" aria-keyshortcuts="/ s" aria-controls="searchbar">
|
|
<i class="fa fa-search"></i>
|
|
</button>
|
|
</div>
|
|
|
|
<h1 class="menu-title">VAPORA Platform Documentation</h1>
|
|
|
|
<div class="right-buttons">
|
|
<a href="../print.html" title="Print this book" aria-label="Print this book">
|
|
<i id="print-button" class="fa fa-print"></i>
|
|
</a>
|
|
<a href="https://github.com/vapora-platform/vapora" title="Git repository" aria-label="Git repository">
|
|
<i id="git-repository-button" class="fa fa-github"></i>
|
|
</a>
|
|
<a href="https://github.com/vapora-platform/vapora/edit/main/docs/src/../adrs/0012-llm-routing-tiers.md" title="Suggest an edit" aria-label="Suggest an edit">
|
|
<i id="git-edit-button" class="fa fa-edit"></i>
|
|
</a>
|
|
|
|
</div>
|
|
</div>
|
|
|
|
<div id="search-wrapper" class="hidden">
|
|
<form id="searchbar-outer" class="searchbar-outer">
|
|
<input type="search" id="searchbar" name="searchbar" placeholder="Search this book ..." aria-controls="searchresults-outer" aria-describedby="searchresults-header">
|
|
</form>
|
|
<div id="searchresults-outer" class="searchresults-outer hidden">
|
|
<div id="searchresults-header" class="searchresults-header"></div>
|
|
<ul id="searchresults">
|
|
</ul>
|
|
</div>
|
|
</div>
|
|
|
|
<!-- Apply ARIA attributes after the sidebar and the sidebar toggle button are added to the DOM -->
|
|
<script>
|
|
document.getElementById('sidebar-toggle').setAttribute('aria-expanded', sidebar === 'visible');
|
|
document.getElementById('sidebar').setAttribute('aria-hidden', sidebar !== 'visible');
|
|
Array.from(document.querySelectorAll('#sidebar a')).forEach(function(link) {
|
|
link.setAttribute('tabIndex', sidebar === 'visible' ? 0 : -1);
|
|
});
|
|
</script>
|
|
|
|
<div id="content" class="content">
|
|
<main>
|
|
<h1 id="adr-012-three-tier-llm-routing-rules--dynamic--override"><a class="header" href="#adr-012-three-tier-llm-routing-rules--dynamic--override">ADR-012: Three-Tier LLM Routing (Rules + Dynamic + Override)</a></h1>
|
|
<p><strong>Status</strong>: Accepted | Implemented
|
|
<strong>Date</strong>: 2024-11-01
|
|
<strong>Deciders</strong>: LLM Architecture Team
|
|
<strong>Technical Story</strong>: Balancing predictability (static rules) with flexibility (dynamic selection) in provider routing</p>
|
|
<hr />
|
|
<h2 id="decision"><a class="header" href="#decision">Decision</a></h2>
|
|
<p>Implementar <strong>three-tier routing system</strong> para seleción de LLM providers: Rules → Dynamic → Override.</p>
|
|
<hr />
|
|
<h2 id="rationale"><a class="header" href="#rationale">Rationale</a></h2>
|
|
<ol>
|
|
<li><strong>Rules-Based</strong>: Predictable routing para tareas conocidas (Architecture → Claude Opus)</li>
|
|
<li><strong>Dynamic</strong>: Runtime selection basado en availability, latency, budget</li>
|
|
<li><strong>Override</strong>: Manual selection con audit logging para troubleshooting/testing</li>
|
|
<li><strong>Balance</strong>: Combinación de determinismo y flexibilidad</li>
|
|
</ol>
|
|
<hr />
|
|
<h2 id="alternatives-considered"><a class="header" href="#alternatives-considered">Alternatives Considered</a></h2>
|
|
<h3 id="-static-rules-only"><a class="header" href="#-static-rules-only">❌ Static Rules Only</a></h3>
|
|
<ul>
|
|
<li><strong>Pros</strong>: Predictable, simple</li>
|
|
<li><strong>Cons</strong>: No adaptación a provider failures, no dynamic cost optimization</li>
|
|
</ul>
|
|
<h3 id="-dynamic-only"><a class="header" href="#-dynamic-only">❌ Dynamic Only</a></h3>
|
|
<ul>
|
|
<li><strong>Pros</strong>: Flexible, adapts to runtime conditions</li>
|
|
<li><strong>Cons</strong>: Unpredictable routing, harder to debug, cold-start problem</li>
|
|
</ul>
|
|
<h3 id="-three-tier-hybrid-chosen"><a class="header" href="#-three-tier-hybrid-chosen">✅ Three-Tier Hybrid (CHOSEN)</a></h3>
|
|
<ul>
|
|
<li>Predictable baseline + flexible adaptation + manual override</li>
|
|
</ul>
|
|
<hr />
|
|
<h2 id="trade-offs"><a class="header" href="#trade-offs">Trade-offs</a></h2>
|
|
<p><strong>Pros</strong>:</p>
|
|
<ul>
|
|
<li>✅ Predictable baseline (rules)</li>
|
|
<li>✅ Automatic adaptation (dynamic)</li>
|
|
<li>✅ Manual control when needed (override)</li>
|
|
<li>✅ Audit trail of decisions</li>
|
|
<li>✅ Graceful degradation</li>
|
|
</ul>
|
|
<p><strong>Cons</strong>:</p>
|
|
<ul>
|
|
<li>⚠️ Added complexity (3 selection layers)</li>
|
|
<li>⚠️ Rule configuration maintenance</li>
|
|
<li>⚠️ Override can introduce inconsistency if overused</li>
|
|
</ul>
|
|
<hr />
|
|
<h2 id="implementation"><a class="header" href="#implementation">Implementation</a></h2>
|
|
<p><strong>Tier 1: Rules-Based Routing</strong>:</p>
|
|
<pre><pre class="playground"><code class="language-rust"><span class="boring">#![allow(unused)]
|
|
</span><span class="boring">fn main() {
|
|
</span>// crates/vapora-llm-router/src/router.rs
|
|
pub struct RoutingRules {
|
|
rules: Vec<(Pattern, ProviderId)>,
|
|
}
|
|
|
|
impl RoutingRules {
|
|
pub fn apply(&self, task: &Task) -> Option<ProviderId> {
|
|
for (pattern, provider) in &self.rules {
|
|
if pattern.matches(&task.description) {
|
|
return Some(provider.clone());
|
|
}
|
|
}
|
|
None
|
|
}
|
|
}
|
|
|
|
// Example rules
|
|
let rules = vec![
|
|
(Pattern::contains("architecture"), "claude-opus"),
|
|
(Pattern::contains("code generation"), "gpt-4"),
|
|
(Pattern::contains("quick query"), "gemini-flash"),
|
|
(Pattern::contains("test"), "ollama"),
|
|
];
|
|
<span class="boring">}</span></code></pre></pre>
|
|
<p><strong>Tier 2: Dynamic Selection</strong>:</p>
|
|
<pre><pre class="playground"><code class="language-rust"><span class="boring">#![allow(unused)]
|
|
</span><span class="boring">fn main() {
|
|
</span>pub async fn select_dynamic(
|
|
task: &Task,
|
|
providers: &[LLMClient],
|
|
) -> Result<&LLMClient> {
|
|
// Score providers by: availability, latency, cost
|
|
let scores: Vec<(ProviderId, f64)> = providers
|
|
.iter()
|
|
.map(|p| {
|
|
let availability = check_availability(p).await;
|
|
let latency = estimate_latency(p).await;
|
|
let cost = get_cost_per_token(p);
|
|
|
|
let score = availability * 0.5
|
|
- latency_penalty(latency) * 0.3
|
|
- cost_penalty(cost) * 0.2;
|
|
(p.id.clone(), score)
|
|
})
|
|
.collect();
|
|
|
|
// Select highest scoring provider
|
|
scores
|
|
.into_iter()
|
|
.max_by(|a, b| a.1.partial_cmp(&b.1).unwrap())
|
|
.ok_or(Error::NoProvidersAvailable)
|
|
}
|
|
<span class="boring">}</span></code></pre></pre>
|
|
<p><strong>Tier 3: Manual Override</strong>:</p>
|
|
<pre><pre class="playground"><code class="language-rust"><span class="boring">#![allow(unused)]
|
|
</span><span class="boring">fn main() {
|
|
</span>pub async fn route_task(
|
|
task: &Task,
|
|
override_provider: Option<ProviderId>,
|
|
) -> Result<String> {
|
|
let provider_id = if let Some(override_id) = override_provider {
|
|
// Tier 3: Manual override (log for audit)
|
|
audit_log::log_override(&task.id, &override_id, &current_user())?;
|
|
override_id
|
|
} else if let Some(rule_provider) = apply_routing_rules(task) {
|
|
// Tier 1: Rules-based
|
|
rule_provider
|
|
} else {
|
|
// Tier 2: Dynamic selection
|
|
select_dynamic(task, &self.providers).await?.id.clone()
|
|
};
|
|
|
|
self.clients
|
|
.get(&provider_id)
|
|
.complete(&task.prompt)
|
|
.await
|
|
}
|
|
<span class="boring">}</span></code></pre></pre>
|
|
<p><strong>Configuration</strong>:</p>
|
|
<pre><code class="language-toml"># config/llm-routing.toml
|
|
|
|
# Tier 1: Rules
|
|
[[routing_rules]]
|
|
pattern = "architecture"
|
|
provider = "claude"
|
|
model = "claude-opus"
|
|
|
|
[[routing_rules]]
|
|
pattern = "code_generation"
|
|
provider = "openai"
|
|
model = "gpt-4"
|
|
|
|
[[routing_rules]]
|
|
pattern = "quick_query"
|
|
provider = "gemini"
|
|
model = "gemini-flash"
|
|
|
|
[[routing_rules]]
|
|
pattern = "test"
|
|
provider = "ollama"
|
|
model = "llama2"
|
|
|
|
# Tier 2: Dynamic scoring weights
|
|
[dynamic_scoring]
|
|
availability_weight = 0.5
|
|
latency_weight = 0.3
|
|
cost_weight = 0.2
|
|
|
|
# Tier 3: Override audit settings
|
|
[override_audit]
|
|
log_all_overrides = true
|
|
require_reason = true
|
|
</code></pre>
|
|
<p><strong>Key Files</strong>:</p>
|
|
<ul>
|
|
<li><code>/crates/vapora-llm-router/src/router.rs</code> (routing logic)</li>
|
|
<li><code>/crates/vapora-llm-router/src/config.rs</code> (rule definitions)</li>
|
|
<li><code>/crates/vapora-backend/src/audit.rs</code> (override logging)</li>
|
|
</ul>
|
|
<hr />
|
|
<h2 id="verification"><a class="header" href="#verification">Verification</a></h2>
|
|
<pre><code class="language-bash"># Test rules-based routing
|
|
cargo test -p vapora-llm-router test_rules_routing
|
|
|
|
# Test dynamic scoring
|
|
cargo test -p vapora-llm-router test_dynamic_scoring
|
|
|
|
# Test override with audit logging
|
|
cargo test -p vapora-llm-router test_override_audit
|
|
|
|
# Integration test: task routing through all tiers
|
|
cargo test -p vapora-llm-router test_full_routing_pipeline
|
|
|
|
# Verify audit trail
|
|
cargo run -p vapora-backend -- audit query --type llm_override --limit 50
|
|
</code></pre>
|
|
<p><strong>Expected Output</strong>:</p>
|
|
<ul>
|
|
<li>Rules correctly match task patterns</li>
|
|
<li>Dynamic scoring selects best available provider</li>
|
|
<li>Overrides logged with user and reason</li>
|
|
<li>Fallback to next tier if previous fails</li>
|
|
<li>All three tiers functional and audited</li>
|
|
</ul>
|
|
<hr />
|
|
<h2 id="consequences"><a class="header" href="#consequences">Consequences</a></h2>
|
|
<h3 id="operational"><a class="header" href="#operational">Operational</a></h3>
|
|
<ul>
|
|
<li>Routing rules maintained in Git (versioned)</li>
|
|
<li>Dynamic scoring requires provider health checks</li>
|
|
<li>Overrides tracked in audit trail for compliance</li>
|
|
</ul>
|
|
<h3 id="performance"><a class="header" href="#performance">Performance</a></h3>
|
|
<ul>
|
|
<li>Rule matching: O(n) patterns (pre-compiled for speed)</li>
|
|
<li>Dynamic scoring: Concurrent provider checks (~50ms)</li>
|
|
<li>Override bypasses both: immediate execution</li>
|
|
</ul>
|
|
<h3 id="monitoring"><a class="header" href="#monitoring">Monitoring</a></h3>
|
|
<ul>
|
|
<li>Track which tier was used per request</li>
|
|
<li>Alert if dynamic tier used frequently (rules insufficient)</li>
|
|
<li>Report override usage patterns (identify gaps in rules)</li>
|
|
</ul>
|
|
<h3 id="debugging"><a class="header" href="#debugging">Debugging</a></h3>
|
|
<ul>
|
|
<li>Audit trail shows exact routing decision</li>
|
|
<li>Reason recorded for overrides</li>
|
|
<li>Helps identify rule gaps or misconfiguration</li>
|
|
</ul>
|
|
<hr />
|
|
<h2 id="references"><a class="header" href="#references">References</a></h2>
|
|
<ul>
|
|
<li><code>/crates/vapora-llm-router/src/router.rs</code> (routing implementation)</li>
|
|
<li><code>/crates/vapora-llm-router/src/config.rs</code> (rule configuration)</li>
|
|
<li><code>/crates/vapora-backend/src/audit.rs</code> (audit logging)</li>
|
|
<li>ADR-007 (Multi-Provider LLM)</li>
|
|
<li>ADR-015 (Budget Enforcement)</li>
|
|
</ul>
|
|
<hr />
|
|
<p><strong>Related ADRs</strong>: ADR-007 (Multi-Provider), ADR-015 (Budget), ADR-016 (Cost Efficiency)</p>
|
|
|
|
</main>
|
|
|
|
<nav class="nav-wrapper" aria-label="Page navigation">
|
|
<!-- Mobile navigation buttons -->
|
|
<a rel="prev" href="../../adrs/0011-secretumvault.html" class="mobile-nav-chapters previous" title="Previous chapter" aria-label="Previous chapter" aria-keyshortcuts="Left">
|
|
<i class="fa fa-angle-left"></i>
|
|
</a>
|
|
|
|
<a rel="next prefetch" href="../../adrs/0013-knowledge-graph.html" class="mobile-nav-chapters next" title="Next chapter" aria-label="Next chapter" aria-keyshortcuts="Right">
|
|
<i class="fa fa-angle-right"></i>
|
|
</a>
|
|
|
|
<div style="clear: both"></div>
|
|
</nav>
|
|
</div>
|
|
</div>
|
|
|
|
<nav class="nav-wide-wrapper" aria-label="Page navigation">
|
|
<a rel="prev" href="../../adrs/0011-secretumvault.html" class="nav-chapters previous" title="Previous chapter" aria-label="Previous chapter" aria-keyshortcuts="Left">
|
|
<i class="fa fa-angle-left"></i>
|
|
</a>
|
|
|
|
<a rel="next prefetch" href="../../adrs/0013-knowledge-graph.html" class="nav-chapters next" title="Next chapter" aria-label="Next chapter" aria-keyshortcuts="Right">
|
|
<i class="fa fa-angle-right"></i>
|
|
</a>
|
|
</nav>
|
|
|
|
</div>
|
|
|
|
|
|
|
|
|
|
<script>
|
|
window.playground_copyable = true;
|
|
</script>
|
|
|
|
|
|
<script src="../elasticlunr.min.js"></script>
|
|
<script src="../mark.min.js"></script>
|
|
<script src="../searcher.js"></script>
|
|
|
|
<script src="../clipboard.min.js"></script>
|
|
<script src="../highlight.js"></script>
|
|
<script src="../book.js"></script>
|
|
|
|
<!-- Custom JS scripts -->
|
|
|
|
|
|
</div>
|
|
</body>
|
|
</html>
|