mirror of
https://github.com/QuantumNous/new-api.git
synced 2026-04-19 07:07:27 +00:00
🛡️ fix: prevent OOM on large/decompressed requests; skip heavy prompt meta when token count is disabled
Clamp request body size (including post-decompression) to avoid memory exhaustion caused by huge payloads/zip bombs, especially with large-context Claude requests. Add a configurable `MAX_REQUEST_BODY_MB` (default `32`) and document it. - Enforce max request body size after gzip/br decompression via `http.MaxBytesReader` - Add a secondary size guard in `common.GetRequestBody` and cache-safe handling - Return **413 Request Entity Too Large** on oversized bodies in relay entry - Avoid building large `TokenCountMeta.CombineText` when both token counting and sensitive check are disabled (use lightweight meta for pricing) - Update READMEs (CN/EN/FR/JA) with `MAX_REQUEST_BODY_MB` - Fix a handful of vet/formatting issues encountered during the change - `go test ./...` passes
This commit is contained in:
@@ -162,7 +162,7 @@ func getModelRequest(c *gin.Context) (*ModelRequest, bool, error) {
|
||||
}
|
||||
midjourneyModel, mjErr, success := service.GetMjRequestModel(relayMode, &midjourneyRequest)
|
||||
if mjErr != nil {
|
||||
return nil, false, fmt.Errorf(mjErr.Description)
|
||||
return nil, false, fmt.Errorf("%s", mjErr.Description)
|
||||
}
|
||||
if midjourneyModel == "" {
|
||||
if !success {
|
||||
|
||||
@@ -5,32 +5,69 @@ import (
|
||||
"io"
|
||||
"net/http"
|
||||
|
||||
"github.com/QuantumNous/new-api/constant"
|
||||
"github.com/andybalholm/brotli"
|
||||
"github.com/gin-gonic/gin"
|
||||
)
|
||||
|
||||
type readCloser struct {
|
||||
io.Reader
|
||||
closeFn func() error
|
||||
}
|
||||
|
||||
func (rc *readCloser) Close() error {
|
||||
if rc.closeFn != nil {
|
||||
return rc.closeFn()
|
||||
}
|
||||
return nil
|
||||
}
|
||||
|
||||
func DecompressRequestMiddleware() gin.HandlerFunc {
|
||||
return func(c *gin.Context) {
|
||||
if c.Request.Body == nil || c.Request.Method == http.MethodGet {
|
||||
c.Next()
|
||||
return
|
||||
}
|
||||
maxMB := constant.MaxRequestBodyMB
|
||||
if maxMB <= 0 {
|
||||
maxMB = 64
|
||||
}
|
||||
maxBytes := int64(maxMB) << 20
|
||||
|
||||
origBody := c.Request.Body
|
||||
wrapMaxBytes := func(body io.ReadCloser) io.ReadCloser {
|
||||
return http.MaxBytesReader(c.Writer, body, maxBytes)
|
||||
}
|
||||
|
||||
switch c.GetHeader("Content-Encoding") {
|
||||
case "gzip":
|
||||
gzipReader, err := gzip.NewReader(c.Request.Body)
|
||||
gzipReader, err := gzip.NewReader(origBody)
|
||||
if err != nil {
|
||||
_ = origBody.Close()
|
||||
c.AbortWithStatus(http.StatusBadRequest)
|
||||
return
|
||||
}
|
||||
defer gzipReader.Close()
|
||||
|
||||
// Replace the request body with the decompressed data
|
||||
c.Request.Body = io.NopCloser(gzipReader)
|
||||
// Replace the request body with the decompressed data, and enforce a max size (post-decompression).
|
||||
c.Request.Body = wrapMaxBytes(&readCloser{
|
||||
Reader: gzipReader,
|
||||
closeFn: func() error {
|
||||
_ = gzipReader.Close()
|
||||
return origBody.Close()
|
||||
},
|
||||
})
|
||||
c.Request.Header.Del("Content-Encoding")
|
||||
case "br":
|
||||
reader := brotli.NewReader(c.Request.Body)
|
||||
c.Request.Body = io.NopCloser(reader)
|
||||
reader := brotli.NewReader(origBody)
|
||||
c.Request.Body = wrapMaxBytes(&readCloser{
|
||||
Reader: reader,
|
||||
closeFn: func() error {
|
||||
return origBody.Close()
|
||||
},
|
||||
})
|
||||
c.Request.Header.Del("Content-Encoding")
|
||||
default:
|
||||
// Even for uncompressed bodies, enforce a max size to avoid huge request allocations.
|
||||
c.Request.Body = wrapMaxBytes(origBody)
|
||||
}
|
||||
|
||||
// Continue processing the request
|
||||
|
||||
Reference in New Issue
Block a user