feat(transport): Added configurable compression levels (#2382)

Author: mitsuhiko

sentry_sdk/consts.py

@@ -40,6 +40,7 @@
             "profiles_sample_rate": Optional[float],
             "profiler_mode": Optional[ProfilerMode],
             "otel_powered_performance": Optional[bool],
+            "transport_zlib_compression_level": Optional[int],
         },
         total=False,
     )

sentry_sdk/transport.py

@@ -163,6 +163,11 @@ def __init__(
             proxy_headers=options["proxy_headers"],
         )
 
+        compresslevel = options.get("_experiments", {}).get(
+            "transport_zlib_compression_level"
+        )
+        self._compresslevel = 9 if compresslevel is None else int(compresslevel)
+
         from sentry_sdk import Hub
 
         self.hub_cls = Hub
@@ -338,8 +343,13 @@ def _send_event(
             return None
 
         body = io.BytesIO()
-        with gzip.GzipFile(fileobj=body, mode="w") as f:
-            f.write(json_dumps(event))
+        if self._compresslevel == 0:
+            body.write(json_dumps(event))
+        else:
+            with gzip.GzipFile(
+                fileobj=body, mode="w", compresslevel=self._compresslevel
+            ) as f:
+                f.write(json_dumps(event))
 
         assert self.parsed_dsn is not None
         logger.debug(
@@ -352,10 +362,14 @@ def _send_event(
                 self.parsed_dsn.host,
             )
         )
-        self._send_request(
-            body.getvalue(),
-            headers={"Content-Type": "application/json", "Content-Encoding": "gzip"},
-        )
+
+        headers = {
+            "Content-Type": "application/json",
+        }
+        if self._compresslevel > 0:
+            headers["Content-Encoding"] = "gzip"
+
+        self._send_request(body.getvalue(), headers=headers)
         return None
 
     def _send_envelope(
@@ -390,8 +404,13 @@ def _send_envelope(
             envelope.items.append(client_report_item)
 
         body = io.BytesIO()
-        with gzip.GzipFile(fileobj=body, mode="w") as f:
-            envelope.serialize_into(f)
+        if self._compresslevel == 0:
+            envelope.serialize_into(body)
+        else:
+            with gzip.GzipFile(
+                fileobj=body, mode="w", compresslevel=self._compresslevel
+            ) as f:
+                envelope.serialize_into(f)
 
         assert self.parsed_dsn is not None
         logger.debug(
@@ -401,12 +420,15 @@ def _send_envelope(
             self.parsed_dsn.host,
         )
 
+        headers = {
+            "Content-Type": "application/x-sentry-envelope",
+        }
+        if self._compresslevel > 0:
+            headers["Content-Encoding"] = "gzip"
+
         self._send_request(
             body.getvalue(),
-            headers={
-                "Content-Type": "application/x-sentry-envelope",
-                "Content-Encoding": "gzip",
-            },
+            headers=headers,
             endpoint_type="envelope",
             envelope=envelope,
         )

tests/test_transport.py

@@ -18,7 +18,7 @@
 from sentry_sdk.integrations.logging import LoggingIntegration
 
 
-CapturedData = namedtuple("CapturedData", ["path", "event", "envelope"])
+CapturedData = namedtuple("CapturedData", ["path", "event", "envelope", "compressed"])
 
 
 class CapturingServer(WSGIServer):
@@ -42,15 +42,25 @@ def __call__(self, environ, start_response):
         """
         request = Request(environ)
         event = envelope = None
+        if request.headers.get("content-encoding") == "gzip":
+            rdr = gzip.GzipFile(fileobj=io.BytesIO(request.data))
+            compressed = True
+        else:
+            rdr = io.BytesIO(request.data)
+            compressed = False
+
         if request.mimetype == "application/json":
-            event = parse_json(gzip.GzipFile(fileobj=io.BytesIO(request.data)).read())
+            event = parse_json(rdr.read())
         else:
-            envelope = Envelope.deserialize_from(
-                gzip.GzipFile(fileobj=io.BytesIO(request.data))
-            )
+            envelope = Envelope.deserialize_from(rdr)
 
         self.captured.append(
-            CapturedData(path=request.path, event=event, envelope=envelope)
+            CapturedData(
+                path=request.path,
+                event=event,
+                envelope=envelope,
+                compressed=compressed,
+            )
         )
 
         response = Response(status=self.code)
@@ -81,6 +91,7 @@ def inner(**kwargs):
 @pytest.mark.parametrize("debug", (True, False))
 @pytest.mark.parametrize("client_flush_method", ["close", "flush"])
 @pytest.mark.parametrize("use_pickle", (True, False))
+@pytest.mark.parametrize("compressionlevel", (0, 9))
 def test_transport_works(
     capturing_server,
     request,
@@ -90,10 +101,16 @@ def test_transport_works(
     make_client,
     client_flush_method,
     use_pickle,
+    compressionlevel,
     maybe_monkeypatched_threading,
 ):
     caplog.set_level(logging.DEBUG)
-    client = make_client(debug=debug)
+    client = make_client(
+        debug=debug,
+        _experiments={
+            "transport_zlib_compression_level": compressionlevel,
+        },
+    )
 
     if use_pickle:
         client = pickle.loads(pickle.dumps(client))
@@ -109,6 +126,7 @@ def test_transport_works(
     out, err = capsys.readouterr()
     assert not err and not out
     assert capturing_server.captured
+    assert capturing_server.captured[0].compressed == (compressionlevel > 0)
 
     assert any("Sending event" in record.msg for record in caplog.records) == debug