Skip to main content
Glama
troubleshooting_performance.md.DxY6LJcT.js1.99 kB
import{_ as a,c as o,o as r,a6 as t}from"./chunks/framework.CHl2ywxc.js";const m=JSON.parse('{"title":"Performance Troubleshooting","description":"","frontmatter":{},"headers":[],"relativePath":"troubleshooting/performance.md","filePath":"troubleshooting/performance.md","lastUpdated":1755281500000}'),n={name:"troubleshooting/performance.md"};function i(l,e,s,c,d,h){return r(),o("div",null,e[0]||(e[0]=[t('<h1 id="performance-troubleshooting" tabindex="-1">Performance Troubleshooting <a class="header-anchor" href="#performance-troubleshooting" aria-label="Permalink to &quot;Performance Troubleshooting&quot;">​</a></h1><h2 id="symptoms" tabindex="-1">Symptoms <a class="header-anchor" href="#symptoms" aria-label="Permalink to &quot;Symptoms&quot;">​</a></h2><ul><li>Elevated latency on tool calls</li><li>Increased error rates or timeouts</li></ul><h2 id="checks" tabindex="-1">Checks <a class="header-anchor" href="#checks" aria-label="Permalink to &quot;Checks&quot;">​</a></h2><ul><li>Inspect <code>/metrics</code> (Node) and platform dashboards (Workers, Koyeb)</li><li>Verify backends’ <code>/health</code> and logs</li><li>Confirm load-balancing strategy is appropriate for your topology</li></ul><h2 id="tuning" tabindex="-1">Tuning <a class="header-anchor" href="#tuning" aria-label="Permalink to &quot;Tuning&quot;">​</a></h2><ul><li>Increase <code>retry.maxRetries</code> and <code>baseDelayMs</code> judiciously</li><li>Switch to <code>health</code> strategy and feed health scores when available</li><li>Tighten circuit breaker thresholds to fail fast on unhealthy instances</li></ul><h2 id="environment" tabindex="-1">Environment <a class="header-anchor" href="#environment" aria-label="Permalink to &quot;Environment&quot;">​</a></h2><ul><li>Run Node with adequate CPU/memory; consider horizontal scaling</li><li>Use KV-backed tokens in Workers to reduce token cache misses</li></ul>',9)]))}const f=a(n,[["render",i]]);export{m as __pageData,f as default};

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/Jakedismo/master-mcp-server'

If you have feedback or need assistance with the MCP directory API, please join our Discord server