AlistGo/alist · warning
get size from url failed, status code
Error message
get size from url %s failed, status code: %d
What it means
With head_size enabled, the url_tree driver issues a HEAD request to each file URL to learn Content-Length. If the server answers with a status >= 300 (redirect chains not followed to a 2xx, 403 auth wall, 404, 429), getSizeFromUrl reports the URL and status code. The node is still created (size stays 0) because the caller only logs this error.
Solutions
- Verify the URL with curl -I; fix expiry/signature or use a stable URL.
- Disable head_size and provide explicit sizes in the structure if the origin cannot serve HEAD reliably.
- If the origin rejects HEAD with 405/403 but allows GET, proxy it or precompute sizes; some servers can be configured to allow HEAD.
Example fix
// before: url with expired signature file https://cdn/x.mp4?sig=expired // after: refresh the URL or disable head_size and set size inline file.mp4:1234567: https://cdn/x.mp4
Defensive patterns
Strategy: fallback
Validate before calling
func checkHEADOk(fileURL string) bool {
res, err := base.RestyClient.R().Head(fileURL)
return err == nil && res.StatusCode() < 300
} Try / catch
// in-tree behavior: error is logged, size stays 0 — callers should tolerate size 0
if size, err := getSizeFromUrl(u); err != nil {
log.Warnf("size probe failed (%v); using 0", err)
size = 0
} Prevention
- Verify URLs with curl -I before enabling head_size
- Prefer explicit sizes in the structure for signed/expiring URLs
- Disable head_size for origins that block HEAD or bot-protect paths
When it happens
Trigger: HEAD request to a signed/expired URL (403), a URL behind Cloudflare or bot protection (403/429), a redirect loop, or a typo'd/dead origin (404/5xx) during BuildTree with head_size on.
Common situations: Time-limited signed URLs in the config, origins that reject HEAD (some return 405), or anti-bot layers. Noticeable as files showing 0 size in the mount.
Related errors
- access_token expired: provide a refresh_token together with…
- can't create RangeReadCloser since URL is empty in link
- cannot copy parent dir to child
- cannot move parent dir to child
- chunkNum invalid
AI-assisted analysis of AlistGo/alist@843d9dc814 (2026-08-15).
Data as JSON: /api/errors/01bc8ec3f7ee000c.
Report an issue: GitHub.
Appendix: source
Thrown at drivers/url_tree/util.go:188
return nil, errs.ObjectNotFound
}
return &model.Object{
Name: node.Name,
Size: node.Size,
Modified: time.Unix(node.Modified, 0),
IsFolder: !node.isFile(),
Path: path,
}, nil
}
func getSizeFromUrl(url string) (int64, error) {
res, err := base.RestyClient.R().SetDoNotParseResponse(true).Head(url)
if err != nil {
return 0, err
}
defer res.RawResponse.Body.Close()
if res.StatusCode() >= 300 {
return 0, fmt.Errorf("get size from url %s failed, status code: %d", url, res.StatusCode())
}
size, err := strconv.ParseInt(res.Header().Get("Content-Length"), 10, 64)
if err != nil {
return 0, err
}
return size, nil
}
func StringifyTree(node *Node) string {
sb := strings.Builder{}
if node.Level == -1 {
for i, child := range node.Children {
sb.WriteString(StringifyTree(child))
if i < len(node.Children)-1 {
sb.WriteString("\n")
}
}
return sb.String()View on GitHub (pinned to 843d9dc814)