pleroma/test/pleroma/reverse_proxy_test.exs
Lain Soykaf 24bdd9774a
fix(reverse_proxy): release the HTTP client on every outcome
Three paths obtained a client ref and abandoned it without closing:
the content-length rejection in call/3, the invalid-status clause in
request/4, and the success branch of response/6.

Under Hackney 4 an abandoned ref is a permanent process leak. A conn
that finishes or errors without an explicit stop parks in `closed` or
`connected` and never terminates: closed/3 only arms its stop timer for
pooled conns, `connected` gets no state_timeout when idle_timeout is
infinity, and a non-pooled conn monitors hackney_conn_sup rather than
its caller, so the owner-DOWN cleanup never fires. Each parked process
pins every refc binary it read, which for proxied video is tens of MB.

Releasing on every outcome means the gun client must stop releasing in
stream_body/1, or the ref would be removed twice.

Hackney does reclaim a pooled connection itself for a response that
cannot carry a body (204, 304, HEAD), so those refs are marked and left
alone; stopping one would evict a connection another request may
already hold. The mark is gated on the request actually being pooled,
since a direct or proxy-tunnelled conn is never checked in and leaks
unless we stop it.
2026-07-25 18:46:43 +04:00

586 lines
18 KiB
Elixir

# Pleroma: A lightweight social networking server
# Copyright © 2017-2022 Pleroma Authors <https://pleroma.social/>
# SPDX-License-Identifier: AGPL-3.0-only
defmodule Pleroma.ReverseProxyTest do
use Pleroma.Web.ConnCase
import ExUnit.CaptureLog
import Mox
alias Pleroma.ReverseProxy
alias Pleroma.ReverseProxy.ClientMock
alias Plug.Conn
setup_all do
{:ok, _} = Registry.start_link(keys: :unique, name: ClientMock)
:ok
end
setup :verify_on_exit!
# Every path that obtains a client ref releases it, so tests that do not assert
# on the release still need close/1 to be callable. Recording each call lets
# the tests below pin *exactly* once: a Mox `expect` would silently fall
# through to this stub on a second call.
setup do
test_pid = self()
stub(ClientMock, :close, fn client ->
send(test_pid, {:closed, client})
:ok
end)
:ok
end
defp request_mock(invokes) do
ClientMock
|> expect(:request, fn :get, url, headers, _body, _opts ->
Registry.register(ClientMock, url, 0)
body = headers |> Enum.into(%{}) |> Jason.encode!()
{:ok, 200,
[
{"content-type", "application/json"},
{"content-length", byte_size(body) |> to_string()}
], %{url: url, body: body}}
end)
|> expect(:stream_body, invokes, fn %{url: url, body: body} = client ->
case Registry.lookup(ClientMock, url) do
[{_, 0}] ->
Registry.update_value(ClientMock, url, &(&1 + 1))
{:ok, body, client}
[{_, 1}] ->
Registry.unregister(ClientMock, url)
:done
end
end)
end
describe "reverse proxy" do
test "do not track successful request", %{conn: conn} do
request_mock(2)
url = "/success"
conn = ReverseProxy.call(conn, url)
assert conn.status == 200
assert Cachex.get(:failed_proxy_url_cache, url) == {:ok, nil}
end
test "uses max_read_duration as the default HTTP receive timeout", %{conn: conn} do
ClientMock
|> expect(:request, fn :get, "/timeout", _, _, opts ->
assert opts[:recv_timeout] == 123
{:error, :timeout}
end)
capture_log(fn -> ReverseProxy.call(conn, "/timeout", max_read_duration: 123) end)
end
test "preserves an explicit HTTP receive timeout", %{conn: conn} do
ClientMock
|> expect(:request, fn :get, "/timeout", _, _, opts ->
assert opts[:recv_timeout] == 456
{:error, :timeout}
end)
capture_log(fn ->
ReverseProxy.call(conn, "/timeout", max_read_duration: 123, http: [recv_timeout: 456])
end)
end
test "keeps unlimited read durations unlimited", %{conn: conn} do
for {max_read_duration, url} <- [{0, "/zero-timeout"}, {:infinity, "/infinite-timeout"}] do
ClientMock
|> expect(:request, fn :get, ^url, _, _, opts ->
assert opts[:recv_timeout] == :infinity
{:error, :timeout}
end)
capture_log(fn ->
ReverseProxy.call(conn, url, max_read_duration: max_read_duration)
end)
end
end
end
# The connection behind a client ref is not tied to the lifetime of this
# request. Hackney 4 in particular parks a non-pooled `hackney_conn` process
# in the `closed` state forever when nobody stops it, pinning every refc
# binary it read, so abandoning a ref leaks the whole response body.
describe "client ref release" do
test "closes the client after the body is fully streamed", %{conn: conn} do
ClientMock
|> expect(:request, fn :get, "/release", _, _, _ ->
{:ok, 200, [{"content-type", "image/png"}], %{url: "/release"}}
end)
|> expect(:stream_body, fn _ -> :done end)
conn = ReverseProxy.call(conn, "/release")
assert conn.halted
assert_received {:closed, %{url: "/release"}}
refute_received {:closed, _}
end
test "closes the client when content-length exceeds max_body_length", %{conn: conn} do
ClientMock
|> expect(:request, fn :get, "/release-huge", _, _, _ ->
{:ok, 200, [{"content-length", "102227307"}], %{url: "/release-huge"}}
end)
assert capture_log(fn ->
ReverseProxy.call(conn, "/release-huge", max_body_length: 25 * 1_048_576)
end) =~ ":body_too_large"
assert_received {:closed, %{url: "/release-huge"}}
refute_received {:closed, _}
end
test "closes the client when the upstream status is not proxiable", %{conn: conn} do
ClientMock
|> expect(:request, fn :get, "/release-418", _, _, _ ->
{:ok, 418, [], %{url: "/release-418"}}
end)
assert capture_log(fn -> ReverseProxy.call(conn, "/release-418") end) =~
"failed with HTTP status 418"
assert_received {:closed, %{url: "/release-418"}}
refute_received {:closed, _}
end
# The hackney client answers HEAD with a ref too, so this path must release
# it as well; only the ref-less 3-tuple reply has nothing to close.
test "closes the client when a HEAD response comes back with a ref", %{conn: conn} do
ClientMock
|> expect(:request, fn :head, "/release-head", _, _, _ ->
{:ok, 200, [{"content-type", "image/png"}], %{url: "/release-head"}}
end)
|> expect(:stream_body, fn _ -> :done end)
conn = ReverseProxy.call(Map.put(conn, :method, "HEAD"), "/release-head")
assert conn.halted
assert_received {:closed, %{url: "/release-head"}}
refute_received {:closed, _}
end
test "has nothing to close when the reply carries no ref", %{conn: conn} do
ClientMock
|> expect(:request, fn :head, "/release-headless", _, _, _ ->
{:ok, 200, [{"content-type", "image/png"}]}
end)
conn = ReverseProxy.call(Map.put(conn, :method, "HEAD"), "/release-headless")
assert conn.status == 200
refute_received {:closed, _}
end
end
test "use Pleroma's user agent in the request; don't pass the client's", %{conn: conn} do
request_mock(2)
conn =
conn
|> Plug.Conn.put_req_header("user-agent", "fake/1.0")
|> ReverseProxy.call("/user-agent")
# Convert the response to a map without relying on json_response
body = conn.resp_body
assert conn.status == 200
response = Jason.decode!(body)
assert response == %{"user-agent" => Pleroma.Application.user_agent()}
end
test "closed connection", %{conn: conn} do
ClientMock
|> expect(:request, fn :get, "/closed", _, _, _ -> {:ok, 200, [], %{}} end)
|> expect(:stream_body, fn _ -> {:error, :closed} end)
|> expect(:close, fn _ -> :ok end)
conn = ReverseProxy.call(conn, "/closed")
assert conn.halted
end
defp stream_mock(invokes) do
ClientMock
|> expect(:request, fn :get, "/stream-bytes/" <> length, _, _, _ ->
Registry.register(ClientMock, "/stream-bytes/" <> length, 0)
{:ok, 200, [{"content-type", "application/octet-stream"}],
%{url: "/stream-bytes/" <> length}}
end)
|> expect(:stream_body, invokes, fn %{url: "/stream-bytes/" <> length} = client ->
max = String.to_integer(length)
case Registry.lookup(ClientMock, "/stream-bytes/" <> length) do
[{_, current}] when current < max ->
Registry.update_value(
ClientMock,
"/stream-bytes/" <> length,
&(&1 + 10)
)
{:ok, "0123456789", client}
[{_, ^max}] ->
Registry.unregister(ClientMock, "/stream-bytes/" <> length)
:done
end
end)
end
describe "max_body" do
test "length returns error if content-length more than option", %{conn: conn} do
request_mock(0)
assert capture_log(fn ->
ReverseProxy.call(conn, "/huge-file", max_body_length: 4)
end) =~
"[error] Elixir.Pleroma.ReverseProxy: request to \"/huge-file\" failed: :body_too_large"
assert {:ok, true} == Cachex.get(:failed_proxy_url_cache, "/huge-file")
assert capture_log(fn ->
ReverseProxy.call(conn, "/huge-file", max_body_length: 4)
end) == ""
end
test "max_body_length returns error if streaming body more than that option", %{conn: conn} do
stream_mock(3)
assert capture_log(fn ->
ReverseProxy.call(conn, "/stream-bytes/50", max_body_length: 30)
end) =~
"Elixir.Pleroma.ReverseProxy request to /stream-bytes/50 failed while reading/chunking: :body_too_large"
assert_received {:closed, %{url: "/stream-bytes/50"}}
refute_received {:closed, _}
end
end
describe "HEAD requests" do
test "common", %{conn: conn} do
ClientMock
|> expect(:request, fn :head, "/head", _, _, _ ->
{:ok, 200, [{"content-type", "image/png"}]}
end)
conn = ReverseProxy.call(Map.put(conn, :method, "HEAD"), "/head")
assert conn.status == 200
assert Conn.get_resp_header(conn, "content-type") == ["image/png"]
assert conn.resp_body == ""
end
end
defp error_mock(status) when is_integer(status) do
ClientMock
|> expect(:request, fn :get, "/status/" <> _, _, _, _ ->
{:error, status}
end)
end
describe "returns error on" do
test "500", %{conn: conn} do
error_mock(500)
url = "/status/500"
capture_log(fn -> ReverseProxy.call(conn, url) end) =~
"[error] Elixir.Pleroma.ReverseProxy: request to /status/500 failed with HTTP status 500"
assert Cachex.get(:failed_proxy_url_cache, url) == {:ok, true}
{:ok, ttl} = Cachex.ttl(:failed_proxy_url_cache, url)
assert ttl <= 60_000
end
test "400", %{conn: conn} do
error_mock(400)
url = "/status/400"
capture_log(fn -> ReverseProxy.call(conn, url) end) =~
"[error] Elixir.Pleroma.ReverseProxy: request to /status/400 failed with HTTP status 400"
assert Cachex.get(:failed_proxy_url_cache, url) == {:ok, true}
assert Cachex.ttl(:failed_proxy_url_cache, url) == {:ok, nil}
end
test "403", %{conn: conn} do
error_mock(403)
url = "/status/403"
capture_log(fn ->
ReverseProxy.call(conn, url, failed_request_ttl: :timer.seconds(120))
end) =~
"[error] Elixir.Pleroma.ReverseProxy: request to /status/403 failed with HTTP status 403"
{:ok, ttl} = Cachex.ttl(:failed_proxy_url_cache, url)
assert ttl > 100_000
end
test "204", %{conn: conn} do
url = "/status/204"
expect(ClientMock, :request, fn :get, _url, _, _, _ -> {:ok, 204, [], %{}} end)
capture_log(fn ->
conn = ReverseProxy.call(conn, url)
assert conn.resp_body == "Request failed: No Content"
assert conn.halted
end) =~
"[error] Elixir.Pleroma.ReverseProxy: request to \"/status/204\" failed with HTTP status 204"
assert Cachex.get(:failed_proxy_url_cache, url) == {:ok, true}
assert Cachex.ttl(:failed_proxy_url_cache, url) == {:ok, nil}
end
end
test "streaming", %{conn: conn} do
stream_mock(21)
conn = ReverseProxy.call(conn, "/stream-bytes/200")
assert conn.state == :chunked
assert byte_size(conn.resp_body) == 200
assert Conn.get_resp_header(conn, "content-type") == ["application/octet-stream"]
end
defp headers_mock(_) do
ClientMock
|> expect(:request, fn :get, "/headers", headers, _, _ ->
Registry.register(ClientMock, "/headers", 0)
{:ok, 200, [{"content-type", "application/json"}], %{url: "/headers", headers: headers}}
end)
|> expect(:stream_body, 2, fn %{url: url, headers: headers} = client ->
case Registry.lookup(ClientMock, url) do
[{_, 0}] ->
Registry.update_value(ClientMock, url, &(&1 + 1))
headers = for {k, v} <- headers, into: %{}, do: {String.capitalize(k), v}
{:ok, Jason.encode!(%{headers: headers}), client}
[{_, 1}] ->
Registry.unregister(ClientMock, url)
:done
end
end)
:ok
end
describe "keep request headers" do
setup [:headers_mock]
test "header passes", %{conn: conn} do
conn =
Conn.put_req_header(
conn,
"accept",
"text/html"
)
|> ReverseProxy.call("/headers")
body = conn.resp_body
assert conn.status == 200
response = Jason.decode!(body)
headers = response["headers"]
assert headers["Accept"] == "text/html"
end
test "header is filtered", %{conn: conn} do
conn =
Conn.put_req_header(
conn,
"accept-language",
"en-US"
)
|> ReverseProxy.call("/headers")
body = conn.resp_body
assert conn.status == 200
response = Jason.decode!(body)
headers = response["headers"]
refute headers["Accept-Language"]
end
end
test "returns 400 on non GET, HEAD requests", %{conn: conn} do
conn = ReverseProxy.call(Map.put(conn, :method, "POST"), "/ip")
assert conn.status == 400
end
describe "cache resp headers" do
test "add cache-control", %{conn: conn} do
ClientMock
|> expect(:request, fn :get, "/cache", _, _, _ ->
{:ok, 200, [{"ETag", "some ETag"}], %{}}
end)
|> expect(:stream_body, fn _ -> :done end)
conn = ReverseProxy.call(conn, "/cache")
assert {"cache-control", "public, max-age=1209600, immutable"} in conn.resp_headers
end
end
defp disposition_headers_mock(headers) do
ClientMock
|> expect(:request, fn :get, "/disposition", _, _, _ ->
Registry.register(ClientMock, "/disposition", 0)
{:ok, 200, headers, %{url: "/disposition"}}
end)
|> expect(:stream_body, 2, fn %{url: "/disposition"} = client ->
case Registry.lookup(ClientMock, "/disposition") do
[{_, 0}] ->
Registry.update_value(ClientMock, "/disposition", &(&1 + 1))
{:ok, "", client}
[{_, 1}] ->
Registry.unregister(ClientMock, "/disposition")
:done
end
end)
end
describe "response content disposition header" do
test "not attachment", %{conn: conn} do
disposition_headers_mock([
{"content-type", "image/gif"},
{"content-length", "0"}
])
conn = ReverseProxy.call(conn, "/disposition")
assert {"content-type", "image/gif"} in conn.resp_headers
end
test "with content-disposition header", %{conn: conn} do
disposition_headers_mock([
{"content-disposition", "attachment; filename=\"filename.jpg\""},
{"content-length", "0"}
])
conn = ReverseProxy.call(conn, "/disposition")
assert {"content-disposition", "attachment; filename=\"filename.jpg\""} in conn.resp_headers
end
end
describe "content-type sanitisation" do
test "preserves allowed image type", %{conn: conn} do
ClientMock
|> expect(:request, fn :get, "/content", _, _, _ ->
{:ok, 200, [{"content-type", "image/png"}], %{url: "/content"}}
end)
|> expect(:stream_body, fn _ -> :done end)
conn = ReverseProxy.call(conn, "/content")
assert conn.status == 200
assert Conn.get_resp_header(conn, "content-type") == ["image/png"]
end
test "preserves allowed video type", %{conn: conn} do
ClientMock
|> expect(:request, fn :get, "/content", _, _, _ ->
{:ok, 200, [{"content-type", "video/mp4"}], %{url: "/content"}}
end)
|> expect(:stream_body, fn _ -> :done end)
conn = ReverseProxy.call(conn, "/content")
assert conn.status == 200
assert Conn.get_resp_header(conn, "content-type") == ["video/mp4"]
end
test "sanitizes ActivityPub content type", %{conn: conn} do
ClientMock
|> expect(:request, fn :get, "/content", _, _, _ ->
{:ok, 200, [{"content-type", "application/activity+json"}], %{url: "/content"}}
end)
|> expect(:stream_body, fn _ -> :done end)
conn = ReverseProxy.call(conn, "/content")
assert conn.status == 200
assert Conn.get_resp_header(conn, "content-type") == ["application/octet-stream"]
end
test "sanitizes LD-JSON content type", %{conn: conn} do
ClientMock
|> expect(:request, fn :get, "/content", _, _, _ ->
{:ok, 200, [{"content-type", "application/ld+json"}], %{url: "/content"}}
end)
|> expect(:stream_body, fn _ -> :done end)
conn = ReverseProxy.call(conn, "/content")
assert conn.status == 200
assert Conn.get_resp_header(conn, "content-type") == ["application/octet-stream"]
end
end
# Hackney is used for Reverse Proxy when Hackney or Finch is the Tesla Adapter
# Gun is able to proxy through Tesla, so it does not need testing as the
# test cases in the Pleroma.HTTPTest module are sufficient
describe "Hackney URL encoding:" do
setup do
ClientMock
|> expect(:request, fn
:get,
"https://example.com/emoji/Pack%201/koronebless.png?foo=bar+baz",
_headers,
_body,
_opts ->
{:ok, 200, [{"content-type", "image/png"}], "It works!"}
:get,
"https://example.com/media/foo/bar%20!$&'()*+,;=/:%20@a%20%5Bbaz%5D.mp4",
_headers,
_body,
_opts ->
{:ok, 200, [{"content-type", "video/mp4"}], "Allowed reserved chars."}
:get, "https://example.com/media/unicode%20%F0%9F%99%82%20.gif", _headers, _body, _opts ->
{:ok, 200, [{"content-type", "image/gif"}], "Unicode emoji in path"}
end)
|> stub(:stream_body, fn _ -> :done end)
|> stub(:close, fn _ -> :ok end)
:ok
end
test "properly encodes URLs with spaces", %{conn: conn} do
url_with_space = "https://example.com/emoji/Pack 1/koronebless.png?foo=bar baz"
result = ReverseProxy.call(conn, url_with_space)
assert result.status == 200
end
test "properly encoded URL should not be altered", %{conn: conn} do
properly_encoded_url = "https://example.com/emoji/Pack%201/koronebless.png?foo=bar+baz"
result = ReverseProxy.call(conn, properly_encoded_url)
assert result.status == 200
end
test "properly encodes URLs with allowed reserved characters", %{conn: conn} do
url_with_reserved_chars = "https://example.com/media/foo/bar !$&'()*+,;=/: @a [baz].mp4"
result = ReverseProxy.call(conn, url_with_reserved_chars)
assert result.status == 200
end
test "properly encodes URLs with unicode in path", %{conn: conn} do
url_with_unicode = "https://example.com/media/unicode 🙂 .gif"
result = ReverseProxy.call(conn, url_with_unicode)
assert result.status == 200
end
end
end