CVE-2026-126071: Rails Active Storage: thread exhaustion DoS via streaming

### Impact When Active Storage serves a file through its proxy controllers, `ActiveStorage::Streaming#send_blob_stream` handles an error from the storage service by calling `head` to send a 404 or 500 response. If the error occurs after the first chunk has been written, the response is already committed, and `head` replaces the response body while the server is still reading the original `ActionController::Live` buffer. That buffer is never closed, so the server thread handling the request blocks until the process restarts. Repeated failed downloads could exhaust the server's threads, possibly resulting in a DoS vulnerability. Applications are affected when they store files on a service that streams a download in several requests, such as Amazon S3, an S3-compatible service or Google Cloud Storage, and draw Active Storage's default routes. The proxy routes are drawn whether or not proxy mode is the default for the application. A file deleted while it is being served, or any other storage error partway through a download, triggers the condition. ### Details A storage error raised before the response is committed still produces a 404 or 500 response. After the fix, a storage error raised after the response is committed is re-raised instead, so the server closes the streamed response and releases the thread; the client receives a truncated body, and the exception reaches the application's error reporting. ### Releases The fixed releases are available at the normal locations. Patches for those who cannot upgrade at once: - 8.1: https://github.com/rails/rails/commit/3c67a693fb21870ecb1222e3da59d5c491a62009.patch - 8.0: https://github.com/rails/rails/commit/655b4442de136a3701334607f634d698bf9d5b82.patch ### Credit This issue was found by the Anthropic OSS Scanner (ANT-2026-FMPUMYTB).

An authenticated low-privilege user (or an attacker who can influence file lifecycle) repeatedly requests large files served through the proxy route from S3 or GCS. While each streamed download is in progress, the file is deleted or the storage service returns an error. After the first chunk has been written, the error handler calls `head`, the stream buffer is never closed and the server thread blocks. Repeating this exhausts the thread pool and the application, including any ML data or model portal behind it, stops responding until restart.