Add body_size_limit option to http module

This option limits the maximum body length that will be read from the
HTTP server. It's meant to prevent misconfigured servers from causing
the probe to use too many resources, even if temporarily. It's not an
additional check on the response, for that, use the resulting metrics
(probe_http_content_length, probe_http_uncompressed_body_length, etc).

Signed-off-by: Marcelo E. Magallon <marcelo.magallon@grafana.com>

CONFIGURATION.md

@@ -49,6 +49,14 @@ The other placeholders are specified separately.
   headers:
     [ <string>: <string> ... ]
 
+  # The maximum uncompressed body length in bytes that will be processed. A value of 0 means no limit.
+  #
+  # If the response includes a Content-Length header, it is NOT validated against this value. This
+  # setting is only meant to limit the amount of data that you are willing to read from the server.
+  #
+  # Example: 10MB
+  [ body_size_limit: <size> | default = 0 ]
+
   # The compression algorithm to use to decompress the response (gzip, br, deflate, identity).
   #
   # If an "Accept-Encoding" header is specified, it MUST be such that the compression algorithm

config/config.go

@@ -28,6 +28,7 @@ import (
 
 	yaml "gopkg.in/yaml.v3"
 
+	"github.com/alecthomas/units"
 	"github.com/go-kit/kit/log"
 	"github.com/go-kit/kit/log/level"
 	"github.com/miekg/dns"
@@ -207,6 +208,7 @@ type HTTPProbe struct {
 	Body                         string                  `yaml:"body,omitempty"`
 	HTTPClientConfig             config.HTTPClientConfig `yaml:"http_client_config,inline"`
 	Compression                  string                  `yaml:"compression,omitempty"`
+	BodySizeLimit                units.Base2Bytes        `yaml:"body_size_limit,omitempty"`
 }
 
 type HeaderMatch struct {
@@ -287,6 +289,11 @@ func (s *HTTPProbe) UnmarshalYAML(unmarshal func(interface{}) error) error {
 	if err := unmarshal((*plain)(s)); err != nil {
 		return err
 	}
+
+	if s.BodySizeLimit < 0 {
+		s.BodySizeLimit = math.MaxInt64
+	}
+
 	if err := s.HTTPClientConfig.Validate(); err != nil {
 		return err
 	}

config/testdata/blackbox-good.yml

@@ -11,6 +11,7 @@ modules:
       basic_auth:
         username: "username"
         password: "mysecret"
+      body_size_limit: 1MB
   tcp_connect:
     prober: tcp
     timeout: 5s

go.mod

@@ -1,6 +1,7 @@
 module github.com/prometheus/blackbox_exporter
 
 require (
+	github.com/alecthomas/units v0.0.0-20210912230133-d1bdfacee922
 	github.com/andybalholm/brotli v1.0.2
 	github.com/go-kit/kit v0.10.0
 	github.com/miekg/dns v1.1.41

go.sum

@@ -13,6 +13,8 @@ github.com/alecthomas/units v0.0.0-20151022065526-2efee857e7cf/go.mod h1:ybxpYRF
 github.com/alecthomas/units v0.0.0-20190717042225-c3de453c63f4/go.mod h1:ybxpYRFXyAe+OPACYpWeL0wqObRcbAqCMya13uyzqw0=
 github.com/alecthomas/units v0.0.0-20190924025748-f65c72e2690d h1:UQZhZ2O0vMHr2cI+DC1Mbh0TJxzA3RcLoMsFw+aXw7E=
 github.com/alecthomas/units v0.0.0-20190924025748-f65c72e2690d/go.mod h1:rBZYJk541a8SKzHPHnH3zbiI+7dagKZ0cgpgrD7Fyho=
+github.com/alecthomas/units v0.0.0-20210912230133-d1bdfacee922 h1:8ypNbf5sd3Sm3cKJ9waOGoQv6dKAFiFty9L6NP1AqJ4=
+github.com/alecthomas/units v0.0.0-20210912230133-d1bdfacee922/go.mod h1:OMCwj8VM1Kc9e19TLln2VL61YJF0x1XFtfdL4JdbSyE=
 github.com/andybalholm/brotli v1.0.2 h1:JKnhI/XQ75uFBTiuzXpzFrUriDPiZjlOSzh6wXogP0E=
 github.com/andybalholm/brotli v1.0.2/go.mod h1:loMXtMfwqflxFJPmdbJO0a3KNoPuLBgiu3qAvBg8x/Y=
 github.com/apache/thrift v0.12.0/go.mod h1:cp2SuWMxlEZw2r+iP2GNCdIi4C1qmUzdZFSVb+bacwQ=

prober/http.go

@@ -499,6 +499,14 @@ func ProbeHTTP(ctx context.Context, target string, module config.Module, registr
 			}
 		}
 
+		// If there's a configured body_size_limit, wrap the body in the response in a http.MaxBytesReader.
+		// This will read up to BodySizeLimit bytes from the body, and return an error if the response is
+		// larger. It forwards the Close call to the original resp.Body to make sure the TCP connection is
+		// correctly shut down. The limit is applied _after decompression_ if applicable.
+		if httpConfig.BodySizeLimit > 0 {
+			resp.Body = http.MaxBytesReader(nil, resp.Body, int64(httpConfig.BodySizeLimit))
+		}
+
 		byteCounter := &byteCounter{ReadCloser: resp.Body}
 
 		if success && (len(httpConfig.FailIfBodyMatchesRegexp) > 0 || len(httpConfig.FailIfBodyNotMatchesRegexp) > 0) {

prober/http_test.go

@@ -513,6 +513,126 @@ func TestHandlingOfCompressionSetting(t *testing.T) {
 	}
 }
 
+func TestMaxResponseLength(t *testing.T) {
+	const max = 128
+
+	var shortGzippedPayload bytes.Buffer
+	enc := gzip.NewWriter(&shortGzippedPayload)
+	enc.Write(bytes.Repeat([]byte{'A'}, max-1))
+	enc.Close()
+
+	var longGzippedPayload bytes.Buffer
+	enc = gzip.NewWriter(&longGzippedPayload)
+	enc.Write(bytes.Repeat([]byte{'A'}, max+1))
+	enc.Close()
+
+	testcases := map[string]struct {
+		target          string
+		compression     string
+		expectedMetrics map[string]float64
+		expectFailure   bool
+	}{
+		"short": {
+			target: "/short",
+			expectedMetrics: map[string]float64{
+				"probe_http_uncompressed_body_length": float64(max - 1),
+				"probe_http_content_length":           float64(max - 1),
+			},
+		},
+		"long": {
+			target:        "/long",
+			expectFailure: true,
+			expectedMetrics: map[string]float64{
+				"probe_http_content_length": float64(max + 1),
+			},
+		},
+		"short compressed": {
+			target:      "/short-compressed",
+			compression: "gzip",
+			expectedMetrics: map[string]float64{
+				"probe_http_content_length":           float64(shortGzippedPayload.Len()),
+				"probe_http_uncompressed_body_length": float64(max - 1),
+			},
+		},
+		"long compressed": {
+			target:        "/long-compressed",
+			compression:   "gzip",
+			expectFailure: true,
+			expectedMetrics: map[string]float64{
+				"probe_http_content_length":           float64(longGzippedPayload.Len()),
+				"probe_http_uncompressed_body_length": max, // it should stop decompressing at max bytes
+			},
+		},
+	}
+
+	ts := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		var resp []byte
+
+		switch r.URL.Path {
+		case "/short-compressed":
+			resp = shortGzippedPayload.Bytes()
+			w.Header().Add("Content-Encoding", "gzip")
+
+		case "/long-compressed":
+			resp = longGzippedPayload.Bytes()
+			w.Header().Add("Content-Encoding", "gzip")
+
+		case "/long":
+			resp = bytes.Repeat([]byte{'A'}, max+1)
+
+		case "/short":
+			resp = bytes.Repeat([]byte{'A'}, max-1)
+
+		default:
+			w.WriteHeader(http.StatusBadRequest)
+			return
+		}
+
+		w.Header().Set("Content-Length", strconv.Itoa(len(resp)))
+		w.WriteHeader(http.StatusOK)
+		w.Write(resp)
+	}))
+	defer ts.Close()
+
+	for name, tc := range testcases {
+		t.Run(name, func(t *testing.T) {
+			registry := prometheus.NewRegistry()
+			testCTX, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+			defer cancel()
+
+			result := ProbeHTTP(
+				testCTX,
+				ts.URL+tc.target,
+				config.Module{
+					Timeout: time.Second,
+					HTTP: config.HTTPProbe{
+						IPProtocolFallback: true,
+						BodySizeLimit:      max,
+						HTTPClientConfig:   pconfig.DefaultHTTPClientConfig,
+						Compression:        tc.compression,
+					},
+				},
+				registry,
+				log.NewNopLogger(),
+			)
+
+			switch {
+			case tc.expectFailure && result:
+				t.Fatalf("test passed unexpectedly")
+			case !tc.expectFailure && !result:
+				t.Fatalf("test failed unexpectedly")
+			}
+
+			mfs, err := registry.Gather()
+			if err != nil {
+				t.Fatal(err)
+			}
+
+			checkRegistryResults(tc.expectedMetrics, mfs, t)
+		})
+	}
+}
+
 func TestRedirectFollowed(t *testing.T) {
 	ts := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
 		if r.URL.Path == "/" {