Implement a structured logging output system. (#5680)

Author: anoadragon453

.buildkite/docker-compose.py35.pg95.yaml

@@ -6,6 +6,7 @@ services:
     image: postgres:9.5
     environment:
       POSTGRES_PASSWORD: postgres
+    command: -c fsync=off
 
   testenv:
     image: python:3.5

.buildkite/docker-compose.py37.pg11.yaml

@@ -6,6 +6,7 @@ services:
     image: postgres:11
     environment:
       POSTGRES_PASSWORD: postgres
+    command: -c fsync=off
 
   testenv:
     image: python:3.7

.buildkite/docker-compose.py37.pg95.yaml

@@ -6,6 +6,7 @@ services:
     image: postgres:9.5
     environment:
       POSTGRES_PASSWORD: postgres
+    command: -c fsync=off
 
   testenv:
     image: python:3.7

.buildkite/pipeline.yml

@@ -45,6 +45,14 @@ steps:
       - docker#v3.0.1:
           image: "python:3.6"
 
+  - command:
+      - "python -m pip install tox"
+      - "tox -e mypy"
+    label: ":mypy: mypy"
+    plugins:
+      - docker#v3.0.1:
+          image: "python:3.5"
+
   - wait
 
   - command:
@@ -54,6 +62,7 @@ steps:
     label: ":python: 3.5 / SQLite / Old Deps"
     env:
       TRIAL_FLAGS: "-j 2"
+      LANG: "C.UTF-8"
     plugins:
       - docker#v3.0.1:
           image: "ubuntu:xenial"  # We use xenail to get an old sqlite and python

.gitignore

@@ -20,6 +20,7 @@ _trial_temp*/
 /*.signing.key
 /env/
 /homeserver*.yaml
+/logs
 /media_store/
 /uploads
 
@@ -29,13 +30,13 @@ _trial_temp*/
 /.vscode/
 
 # build products
-/.coverage*
 !/.coveragerc
+/.coverage*
+/.mypy_cache/
 /.tox
 /build/
 /coverage.*
 /dist/
 /docs/build/
 /htmlcov
 /pip-wheel-metadata/
-

changelog.d/5680.misc

@@ -0,0 +1 @@
+Lay the groundwork for structured logging output.

docs/structured_logging.md

@@ -0,0 +1,83 @@
+# Structured Logging
+
+A structured logging system can be useful when your logs are destined for a machine to parse and process. By maintaining its machine-readable characteristics, it enables more efficient searching and aggregations when consumed by software such as the "ELK stack".
+
+Synapse's structured logging system is configured via the file that Synapse's `log_config` config option points to. The file must be YAML and contain `structured: true`. It must contain a list of "drains" (places where logs go to).
+
+A structured logging configuration looks similar to the following:
+
+```yaml
+structured: true
+
+loggers:
+    synapse:
+        level: INFO
+    synapse.storage.SQL:
+        level: WARNING
+
+drains:
+    console:
+        type: console
+        location: stdout
+    file:
+        type: file_json
+        location: homeserver.log
+```
+
+The above logging config will set Synapse as 'INFO' logging level by default, with the SQL layer at 'WARNING', and will have two logging drains (to the console and to a file, stored as JSON).
+
+## Drain Types
+
+Drain types can be specified by the `type` key.
+
+### `console`
+
+Outputs human-readable logs to the console.
+
+Arguments:
+
+- `location`: Either `stdout` or `stderr`.
+
+### `console_json`
+
+Outputs machine-readable JSON logs to the console.
+
+Arguments:
+
+- `location`: Either `stdout` or `stderr`.
+
+### `console_json_terse`
+
+Outputs machine-readable JSON logs to the console, separated by newlines. This
+format is not designed to be read and re-formatted into human-readable text, but
+is optimal for a logging aggregation system.
+
+Arguments:
+
+- `location`: Either `stdout` or `stderr`.
+
+### `file`
+
+Outputs human-readable logs to a file.
+
+Arguments:
+
+- `location`: An absolute path to the file to log to.
+
+### `file_json`
+
+Outputs machine-readable logs to a file.
+
+Arguments:
+
+- `location`: An absolute path to the file to log to.
+
+### `network_json_terse`
+
+Delivers machine-readable JSON logs to a log aggregator over TCP. This is
+compatible with LogStash's TCP input with the codec set to `json_lines`.
+
+Arguments:
+
+- `host`: Hostname or IP address of the log aggregator.
+- `port`: Numerical port to contact on the host.

synapse/app/_base.py

@@ -36,18 +36,20 @@
 
 logger = logging.getLogger(__name__)
 
+# list of tuples of function, args list, kwargs dict
 _sighup_callbacks = []
 
 
-def register_sighup(func):
+def register_sighup(func, *args, **kwargs):
     """
     Register a function to be called when a SIGHUP occurs.
 
     Args:
         func (function): Function to be called when sent a SIGHUP signal.
-            Will be called with a single argument, the homeserver.
+            Will be called with a single default argument, the homeserver.
+        *args, **kwargs: args and kwargs to be passed to the target function.
     """
-    _sighup_callbacks.append(func)
+    _sighup_callbacks.append((func, args, kwargs))
 
 
 def start_worker_reactor(appname, config, run_command=reactor.run):
@@ -248,8 +250,8 @@ def handle_sighup(*args, **kwargs):
                 # we're not using systemd.
                 sdnotify(b"RELOADING=1")
 
-                for i in _sighup_callbacks:
-                    i(hs)
+                for i, args, kwargs in _sighup_callbacks:
+                    i(hs, *args, **kwargs)
 
                 sdnotify(b"READY=1")
 

synapse/app/admin_cmd.py

@@ -227,8 +227,6 @@ def start(config_options):
     config.start_pushers = False
     config.send_federation = False
 
-    setup_logging(config, use_worker_options=True)
-
     synapse.events.USE_FROZEN_DICTS = config.use_frozen_dicts
 
     database_engine = create_engine(config.database_config)
@@ -241,6 +239,8 @@ def start(config_options):
         database_engine=database_engine,
     )
 
+    setup_logging(ss, config, use_worker_options=True)
+
     ss.setup()
 
     # We use task.react as the basic run command as it correctly handles tearing

synapse/app/appservice.py

@@ -141,8 +141,6 @@ def start(config_options):
 
     assert config.worker_app == "synapse.app.appservice"
 
-    setup_logging(config, use_worker_options=True)
-
     events.USE_FROZEN_DICTS = config.use_frozen_dicts
 
     database_engine = create_engine(config.database_config)
@@ -167,6 +165,8 @@ def start(config_options):
         database_engine=database_engine,
     )
 
+    setup_logging(ps, config, use_worker_options=True)
+
     ps.setup()
     reactor.addSystemEventTrigger(
         "before", "startup", _base.start, ps, config.worker_listeners