返回 AiToEarn
server-hoist-static-io.md
1 ---
2 title: Hoist Static I/O to Module Level
3 impact: HIGH
4 impactDescription: avoids repeated file/network I/O per request
5 tags: server, io, performance, next.js, route-handlers, og-image
6 ---
7
8 ## Hoist Static I/O to Module Level
9
10 **Impact: HIGH (avoids repeated file/network I/O per request)**
11
12 When loading static assets (fonts, logos, images, config files) in route handlers or server functions, hoist the I/O operation to module level. Module-level code runs once when the module is first imported, not on every request. This eliminates redundant file system reads or network fetches that would otherwise run on every invocation.
13
14 **Incorrect (reads font file on every request):**
15
16 ```typescript
17 // app/api/og/route.tsx
18 import { ImageResponse } from 'next/og'
19
20 export async function GET(request: Request) {
21 // Runs on EVERY request - expensive!
22 const fontData = await fetch(
23 new URL('./fonts/Inter.ttf', import.meta.url)
24 ).then(res => res.arrayBuffer())
25
26 const logoData = await fetch(
27 new URL('./images/logo.png', import.meta.url)
28 ).then(res => res.arrayBuffer())
29
30 return new ImageResponse(
31 <div style={{ fontFamily: 'Inter' }}>
32 <img src={logoData} />
33 Hello World
34 </div>,
35 { fonts: [{ name: 'Inter', data: fontData }] }
36 )
37 }
38 ```
39
40 **Correct (loads once at module initialization):**
41
42 ```typescript
43 // app/api/og/route.tsx
44 import { ImageResponse } from 'next/og'
45
46 // Module-level: runs ONCE when module is first imported
47 const fontData = fetch(
48 new URL('./fonts/Inter.ttf', import.meta.url)
49 ).then(res => res.arrayBuffer())
50
51 const logoData = fetch(
52 new URL('./images/logo.png', import.meta.url)
53 ).then(res => res.arrayBuffer())
54
55 export async function GET(request: Request) {
56 // Await the already-started promises
57 const [font, logo] = await Promise.all([fontData, logoData])
58
59 return new ImageResponse(
60 <div style={{ fontFamily: 'Inter' }}>
61 <img src={logo} />
62 Hello World
63 </div>,
64 { fonts: [{ name: 'Inter', data: font }] }
65 )
66 }
67 ```
68
69 **Correct (synchronous fs at module level):**
70
71 ```typescript
72 // app/api/og/route.tsx
73 import { ImageResponse } from 'next/og'
74 import { readFileSync } from 'fs'
75 import { join } from 'path'
76
77 // Synchronous read at module level - blocks only during module init
78 const fontData = readFileSync(
79 join(process.cwd(), 'public/fonts/Inter.ttf')
80 )
81
82 const logoData = readFileSync(
83 join(process.cwd(), 'public/images/logo.png')
84 )
85
86 export async function GET(request: Request) {
87 return new ImageResponse(
88 <div style={{ fontFamily: 'Inter' }}>
89 <img src={logoData} />
90 Hello World
91 </div>,
92 { fonts: [{ name: 'Inter', data: fontData }] }
93 )
94 }
95 ```
96
97 **Incorrect (reads config on every call):**
98
99 ```typescript
100 import fs from 'node:fs/promises'
101
102 export async function processRequest(data: Data) {
103 const config = JSON.parse(
104 await fs.readFile('./config.json', 'utf-8')
105 )
106 const template = await fs.readFile('./template.html', 'utf-8')
107
108 return render(template, data, config)
109 }
110 ```
111
112 **Correct (hoists config and template to module level):**
113
114 ```typescript
115 import fs from 'node:fs/promises'
116
117 const configPromise = fs
118 .readFile('./config.json', 'utf-8')
119 .then(JSON.parse)
120 const templatePromise = fs.readFile('./template.html', 'utf-8')
121
122 export async function processRequest(data: Data) {
123 const [config, template] = await Promise.all([
124 configPromise,
125 templatePromise,
126 ])
127
128 return render(template, data, config)
129 }
130 ```
131
132 When to use this pattern:
133
134 - Loading fonts for OG image generation
135 - Loading static logos, icons, or watermarks
136 - Reading configuration files that don't change at runtime
137 - Loading email templates or other static templates
138 - Any static asset that's the same across all requests
139
140 When not to use this pattern:
141
142 - Assets that vary per request or user
143 - Files that may change during runtime (use caching with TTL instead)
144 - Large files that would consume too much memory if kept loaded
145 - Sensitive data that shouldn't persist in memory
146
147 With Vercel's [Fluid Compute](https://vercel.com/docs/fluid-compute), module-level caching is especially effective because multiple concurrent requests share the same function instance. The static assets stay loaded in memory across requests without cold start penalties.
148
149 In traditional serverless, each cold start re-executes module-level code, but subsequent warm invocations reuse the loaded assets until the instance is recycled.
150
150 lines MARKDOWN