Allocation of Resources Without Limits or Throttling vulnerability in elixir-grpc grpc allows unauthenticated attackers to exhaust the BEAM's memory and crash the server by streaming a large or slow-trickle unary request body.
'Elixir.GRPC.Server.Adapters.Cowboy.Handler':read_full_body/3 (lib/grpc/server/adapters/cowboy/handler.ex) accumulates every received chunk into a single growing binary with no size cap. Additionally, when the client omits the grpc-timeout header, the per-chunk read timeout resolves to :infinity, allowing a slow-trickle client to keep the connection alive indefinitely while memory grows. A single connection is sufficient to exhaust server memory and crash the node.
This issue affects grpc: from 0.3.0-alpha.2 before 1.0.0.
Affected: Hex: grpc
### Summary
`'Elixir.GRPC.Server.Adapters.Cowboy.Handler':read_full_body/3` accumulates every received chunk into a single growing binary with no size cap. When the client omits the `grpc-timeout` header, the read timeout resolves to `:infinity`, allowing a slow-trickle attacker to hold the connection open indefinitely while memory grows. A single unauthenticated connection is sufficient to exhaust BEAM memory and crash the node.
### Details
The read loop in `lib/grpc/server/adapters/cowboy/handler.ex` calls `:cowboy_req.read_body/2` in a recursive loop, concatenating each chunk: `body <> data`. There is no running-total check and no configurable maximum body size. As the loop drains the receive buffer, cowboy issues fresh HTTP/2 `WINDOW_UPDATE` frames, so the client can keep pushing data indefinitely.
The `grpc-timeout` header is attacker-supplied and optional. When absent, `timeout_left_opt(nil)` returns `:infinity`, so the per-chunk read also has no deadline. The two missing controls compound: a fast client can blast multi-gigabyte payloads directly into memory; a slow client can trickle data forever.
### PoC
1. Start any `grpc` server exposing a unary RPC (no special configuration required). 2. Open an HTTP/2 connection and send a POST to any unary RPC path with `Content-Type: application/grpc+proto` — omit the `grpc-timeout` header. 3. Stream a large body (e.g. 1 GiB) in chunks without sending the final `END_STREAM` flag immediately. 4. Observe BEAM memory growing…