configs: moved all gunicorn config to python file

configs/development.ini

@@ -1,275 +1,204 b''
 #
 
 ; #################################
 ; RHODECODE VCSSERVER CONFIGURATION
 ; #################################
 
 [server:main]
 ; COMMON HOST/IP CONFIG
 host = 0.0.0.0
-port = 9900
+port = 10010
 
 ; ##################################################
 ; WAITRESS WSGI SERVER - Recommended for Development
 ; ##################################################
 
 ; use server type
 use = egg:waitress#main
 
 ; number of worker threads
 threads = 5
 
 ; MAX BODY SIZE 100GB
 max_request_body_size = 107374182400
 
 ; Use poll instead of select, fixes file descriptors limits problems.
 ; May not work on old windows systems.
 asyncore_use_poll = true
 
 
 ; ###########################
 ; GUNICORN APPLICATION SERVER
 ; ###########################
 
 ; run with gunicorn --paste rhodecode.ini
 
 ; Module to use, this setting shouldn't be changed
 #use = egg:gunicorn#main
 
-; Sets the number of process workers. More workers means more concurrent connections
-; RhodeCode can handle at the same time. Each additional worker also it increases
-; memory usage as each has it's own set of caches.
-; Recommended value is (2 * NUMBER_OF_CPUS + 1), eg 2CPU = 5 workers, but no more
-; than 8-10 unless for really big deployments .e.g 700-1000 users.
-; `instance_id = *` must be set in the [app:main] section below (which is the default)
-; when using more than 1 worker.
-#workers = 2
-
-; Gunicorn access log level
-#loglevel = info
-
-; Process name visible in process list
-#proc_name = rhodecode_vcsserver
-
-; Type of worker class, one of `sync`, `gevent`
-; currently `sync` is the only option allowed.
-#worker_class = sync
-
-; The maximum number of simultaneous clients. Valid only for gevent
-#worker_connections = 10
-
-; Max number of requests that worker will handle before being gracefully restarted.
-; Prevents memory leaks, jitter adds variability so not all workers are restarted at once.
-#max_requests = 1000
-#max_requests_jitter = 30
-
-; Amount of time a worker can spend with handling a request before it
-; gets killed and restarted. By default set to 21600 (6hrs)
-; Examples: 1800 (30min), 3600 (1hr), 7200 (2hr), 43200 (12h)
-#timeout = 21600
-
-; The maximum size of HTTP request line in bytes.
-; 0 for unlimited
-#limit_request_line = 0
-
-; Limit the number of HTTP headers fields in a request.
-; By default this value is 100 and can't be larger than 32768.
-#limit_request_fields = 32768
-
-; Limit the allowed size of an HTTP request header field.
-; Value is a positive number or 0.
-; Setting it to 0 will allow unlimited header field sizes.
-#limit_request_field_size = 0
-
-; Timeout for graceful workers restart.
-; After receiving a restart signal, workers have this much time to finish
-; serving requests. Workers still alive after the timeout (starting from the
-; receipt of the restart signal) are force killed.
-; Examples: 1800 (30min), 3600 (1hr), 7200 (2hr), 43200 (12h)
-#graceful_timeout = 21600
-
-# The number of seconds to wait for requests on a Keep-Alive connection.
-# Generally set in the 1-5 seconds range.
-#keepalive = 2
-
-; Maximum memory usage that each worker can use before it will receive a
-; graceful restart signal 0 = memory monitoring is disabled
-; Examples: 268435456 (256MB), 536870912 (512MB)
-; 1073741824 (1GB), 2147483648 (2GB), 4294967296 (4GB)
-#memory_max_usage = 0
-
-; How often in seconds to check for memory usage for each gunicorn worker
-#memory_usage_check_interval = 60
-
-; Threshold value for which we don't recycle worker if GarbageCollection
-; frees up enough resources. Before each restart we try to run GC on worker
-; in case we get enough free memory after that, restart will not happen.
-#memory_usage_recovery_threshold = 0.8
-
-
 [app:main]
 ; The %(here)s variable will be replaced with the absolute path of parent directory
 ; of this file
 ; Each option in the app:main can be override by an environmental variable
 ;
 ;To override an option:
 ;
 ;RC_<KeyName>
 ;Everything should be uppercase, . and - should be replaced by _.
 ;For example, if you have these configuration settings:
 ;rc_cache.repo_object.backend = foo
 ;can be overridden by
 ;export RC_CACHE_REPO_OBJECT_BACKEND=foo
 
 use = egg:rhodecode-vcsserver
 
 
 ; #############
 ; DEBUG OPTIONS
 ; #############
 
 # During development the we want to have the debug toolbar enabled
 pyramid.includes =
     pyramid_debugtoolbar
 
 debugtoolbar.hosts = 0.0.0.0/0
 debugtoolbar.exclude_prefixes =
     /css
     /fonts
     /images
     /js
 
 ; #################
 ; END DEBUG OPTIONS
 ; #################
 
 ; Pyramid default locales, we need this to be set
 #pyramid.default_locale_name = en
 
 ; default locale used by VCS systems
 #locale = en_US.UTF-8
 
 ; path to binaries for vcsserver, it should be set by the installer
 ; at installation time, e.g /home/user/.rccontrol/vcsserver-1/profile/bin
 ; it can also be a path to nix-build output in case of development
 core.binary_dir = ""
 
 ; Custom exception store path, defaults to TMPDIR
 ; This is used to store exception from RhodeCode in shared directory
 #exception_tracker.store_path =
 
 ; #############
 ; DOGPILE CACHE
 ; #############
 
 ; Default cache dir for caches. Putting this into a ramdisk can boost performance.
 ; eg. /tmpfs/data_ramdisk, however this directory might require large amount of space
 #cache_dir = %(here)s/data
 
 ; ***************************************
 ; `repo_object` cache, default file based
 ; ***************************************
 
 ; `repo_object` cache settings for vcs methods for repositories
 #rc_cache.repo_object.backend = dogpile.cache.rc.file_namespace
 
 ; cache auto-expires after N seconds
 ; Examples: 86400 (1Day), 604800 (7Days), 1209600 (14Days), 2592000 (30days), 7776000 (90Days)
 #rc_cache.repo_object.expiration_time = 2592000
 
 ; file cache store path. Defaults to `cache_dir =` value or tempdir if both values are not set
 #rc_cache.repo_object.arguments.filename = /tmp/vcsserver_cache_repo_object.db
 
 ; ***********************************************************
 ; `repo_object` cache with redis backend
 ; recommended for larger instance, and for better performance
 ; ***********************************************************
 
 ; `repo_object` cache settings for vcs methods for repositories
 #rc_cache.repo_object.backend = dogpile.cache.rc.redis_msgpack
 
 ; cache auto-expires after N seconds
 ; Examples: 86400 (1Day), 604800 (7Days), 1209600 (14Days), 2592000 (30days), 7776000 (90Days)
 #rc_cache.repo_object.expiration_time = 2592000
 
 ; redis_expiration_time needs to be greater then expiration_time
 #rc_cache.repo_object.arguments.redis_expiration_time = 3592000
 
 #rc_cache.repo_object.arguments.host = localhost
 #rc_cache.repo_object.arguments.port = 6379
 #rc_cache.repo_object.arguments.db = 5
 #rc_cache.repo_object.arguments.socket_timeout = 30
 ; more Redis options: https://dogpilecache.sqlalchemy.org/en/latest/api.html#redis-backends
 #rc_cache.repo_object.arguments.distributed_lock = true
 
 ; auto-renew lock to prevent stale locks, slower but safer. Use only if problems happen
 #rc_cache.repo_object.arguments.lock_auto_renewal = true
 
 ; Statsd client config, this is used to send metrics to statsd
 ; We recommend setting statsd_exported and scrape them using Promethues
 #statsd.enabled = false
 #statsd.statsd_host = 0.0.0.0
 #statsd.statsd_port = 8125
 #statsd.statsd_prefix =
 #statsd.statsd_ipv6 = false
 
 ; configure logging automatically at server startup set to false
 ; to use the below custom logging config.
 ; RC_LOGGING_FORMATTER
 ; RC_LOGGING_LEVEL
 ; env variables can control the settings for logging in case of autoconfigure
 
 #logging.autoconfigure = true
 
 ; specify your own custom logging config file to configure logging
 #logging.logging_conf_file = /path/to/custom_logging.ini
 
 ; #####################
 ; LOGGING CONFIGURATION
 ; #####################
 
 [loggers]
 keys = root, vcsserver
 
 [handlers]
 keys = console
 
 [formatters]
 keys = generic, json
 
 ; #######
 ; LOGGERS
 ; #######
 [logger_root]
 level = NOTSET
 handlers = console
 
 [logger_vcsserver]
 level = DEBUG
 handlers =
 qualname = vcsserver
 propagate = 1
 
 ; ########
 ; HANDLERS
 ; ########
 
 [handler_console]
 class = StreamHandler
 args = (sys.stderr, )
 level = DEBUG
 ; To enable JSON formatted logs replace 'generic' with 'json'
 ; This allows sending properly formatted logs to grafana loki or elasticsearch
 formatter = generic
 
 ; ##########
 ; FORMATTERS
 ; ##########
 
 [formatter_generic]
 format = %(asctime)s.%(msecs)03d [%(process)d] %(levelname)-5.5s [%(name)s] %(message)s
 datefmt = %Y-%m-%d %H:%M:%S
 
 [formatter_json]
 format = %(timestamp)s %(levelname)s %(name)s %(message)s %(req_id)s
 class = vcsserver.lib._vendor.jsonlogger.JsonFormatter

configs/gunicorn_config.py

@@ -1,393 +1,506 b''
 """
 Gunicorn config extension and hooks. This config file adds some extra settings and memory management.
 Gunicorn configuration should be managed by .ini files entries of RhodeCode or VCSServer
 """
 
 import gc
 import os
 import sys
 import math
 import time
 import threading
 import traceback
 import random
 import socket
+import dataclasses
 from gunicorn.glogging import Logger
 
 
 def get_workers():
     import multiprocessing
     return multiprocessing.cpu_count() * 2 + 1
 
-# GLOBAL
+
+bind = "127.0.0.1:10010"
+
+
+# Error logging output for gunicorn (-) is stdout
 errorlog = '-'
+
+# Access logging output for gunicorn (-) is stdout
 accesslog = '-'
 
 
 # SERVER MECHANICS
 # None == system temp dir
 # worker_tmp_dir is recommended to be set to some tmpfs
 worker_tmp_dir = None
 tmp_upload_dir = None
 
+# use re-use port logic
 #reuse_port = True
 
 # Custom log format
 #access_log_format = (
 #    '%(t)s %(p)s INFO  [GNCRN] %(h)-15s rqt:%(L)s %(s)s %(b)-6s "%(m)s:%(U)s %(q)s" usr:%(u)s "%(f)s" "%(a)s"')
 
 # loki format for easier parsing in grafana
 access_log_format = (
     'time="%(t)s" pid=%(p)s level="INFO" type="[GNCRN]" ip="%(h)-15s" rqt="%(L)s" response_code="%(s)s" response_bytes="%(b)-6s" uri="%(m)s:%(U)s %(q)s" user=":%(u)s" user_agent="%(a)s"')
 
-# self adjust workers based on CPU count
+
+# Sets the number of process workers. More workers means more concurrent connections
+# RhodeCode can handle at the same time. Each additional worker also it increases
+# memory usage as each has it's own set of caches.
+# Recommended value is (2 * NUMBER_OF_CPUS + 1), eg 2CPU = 5 workers, but no more
+# than 8-10 unless for huge deployments .e.g 700-1000 users.
+# `instance_id = *` must be set in the [app:main] section below (which is the default)
+# when using more than 1 worker.
+workers = 6
+
+# self adjust workers based on CPU count, to use maximum of CPU and not overquota the resources
 # workers = get_workers()
 
+# Gunicorn access log level
+loglevel = 'info'
+
+# Process name visible in process list
+proc_name = 'rhodecode_vcsserver'
+
+# Type of worker class, one of `sync`, `gevent`
+# currently `sync` is the only option allowed.
+worker_class = 'sync'
+
+# The maximum number of simultaneous clients. Valid only for gevent
+worker_connections = 10
+
+# Max number of requests that worker will handle before being gracefully restarted.
+# Prevents memory leaks, jitter adds variability so not all workers are restarted at once.
+max_requests = 2000
+max_requests_jitter = 30
+
+# The maximum number of pending connections.
+# Exceeding this number results in the client getting an error when attempting to connect.
+backlog = 64
+
+# Amount of time a worker can spend with handling a request before it
+# gets killed and restarted. By default set to 21600 (6hrs)
+# Examples: 1800 (30min), 3600 (1hr), 7200 (2hr), 43200 (12h)
+timeout = 21600
+
+# The maximum size of HTTP request line in bytes.
+# 0 for unlimited
+limit_request_line = 0
+
+# Limit the number of HTTP headers fields in a request.
+# By default this value is 100 and can't be larger than 32768.
+limit_request_fields = 32768
+
+# Limit the allowed size of an HTTP request header field.
+# Value is a positive number or 0.
+# Setting it to 0 will allow unlimited header field sizes.
+limit_request_field_size = 0
+
+# Timeout for graceful workers restart.
+# After receiving a restart signal, workers have this much time to finish
+# serving requests. Workers still alive after the timeout (starting from the
+# receipt of the restart signal) are force killed.
+# Examples: 1800 (30min), 3600 (1hr), 7200 (2hr), 43200 (12h)
+graceful_timeout = 21600
+
+# The number of seconds to wait for requests on a Keep-Alive connection.
+# Generally set in the 1-5 seconds range.
+keepalive = 2
+
+# Maximum memory usage that each worker can use before it will receive a
+# graceful restart signal 0 = memory monitoring is disabled
+# Examples: 268435456 (256MB), 536870912 (512MB)
+# 1073741824 (1GB), 2147483648 (2GB), 4294967296 (4GB)
+memory_max_usage = 0
+
+# How often in seconds to check for memory usage for each gunicorn worker
+memory_usage_check_interval = 60
+
+# Threshold value for which we don't recycle worker if GarbageCollection
+# frees up enough resources. Before each restart we try to run GC on worker
+# in case we get enough free memory after that, restart will not happen.
+memory_usage_recovery_threshold = 0.8
+
+
[e-mail]
+class MemoryCheckConfig:
+    max_usage: int
+    check_interval: int
+    recovery_threshold: float
+
 
 def _get_process_rss(pid=None):
     try:
         import psutil
         if pid:
             proc = psutil.Process(pid)
         else:
             proc = psutil.Process()
         return proc.memory_info().rss
     except Exception:
         return None
 
 
 def _get_config(ini_path):
     import configparser
 
     try:
         config = configparser.RawConfigParser()
         config.read(ini_path)
         return config
     except Exception:
         return None
 
 
-def _time_with_offset(memory_usage_check_interval):
-    return time.time() - random.randint(0, memory_usage_check_interval/2.0)
+def get_memory_usage_params(config=None):
+    # memory spec defaults
+    _memory_max_usage = memory_max_usage
+    _memory_usage_check_interval = memory_usage_check_interval
+    _memory_usage_recovery_threshold = memory_usage_recovery_threshold
+
+    if config:
+        ini_path = os.path.abspath(config)
+        conf = _get_config(ini_path)
+
+        section = 'server:main'
+        if conf and conf.has_section(section):
+
+            if conf.has_option(section, 'memory_max_usage'):
+                _memory_max_usage = conf.getint(section, 'memory_max_usage')
+
+            if conf.has_option(section, 'memory_usage_check_interval'):
+                _memory_usage_check_interval = conf.getint(section, 'memory_usage_check_interval')
+
+            if conf.has_option(section, 'memory_usage_recovery_threshold'):
+                _memory_usage_recovery_threshold = conf.getfloat(section, 'memory_usage_recovery_threshold')
+
+    _memory_max_usage = int(os.environ.get('RC_GUNICORN_MEMORY_MAX_USAGE', '')
+                            or _memory_max_usage)
+    _memory_usage_check_interval = int(os.environ.get('RC_GUNICORN_MEMORY_USAGE_CHECK_INTERVAL', '')
+                                       or _memory_usage_check_interval)
+    _memory_usage_recovery_threshold = float(os.environ.get('RC_GUNICORN_MEMORY_USAGE_RECOVERY_THRESHOLD', '')
+                                             or _memory_usage_recovery_threshold)
+
+    return MemoryCheckConfig(_memory_max_usage, _memory_usage_check_interval, _memory_usage_recovery_threshold)
+
+
+def _time_with_offset(check_interval):
+    return time.time() - random.randint(0, check_interval/2.0)
 
 
 def pre_fork(server, worker):
     pass
 
 
 def post_fork(server, worker):
 
-    # memory spec defaults
-    _memory_max_usage = 0
-    _memory_usage_check_interval = 60
-    _memory_usage_recovery_threshold = 0.8
-
-    ini_path = os.path.abspath(server.cfg.paste)
-    conf = _get_config(ini_path)
-
-    section = 'server:main'
-    if conf and conf.has_section(section):
-
-        if conf.has_option(section, 'memory_max_usage'):
-            _memory_max_usage = conf.getint(section, 'memory_max_usage')
-
-        if conf.has_option(section, 'memory_usage_check_interval'):
-            _memory_usage_check_interval = conf.getint(section, 'memory_usage_check_interval')
-
-        if conf.has_option(section, 'memory_usage_recovery_threshold'):
-            _memory_usage_recovery_threshold = conf.getfloat(section, 'memory_usage_recovery_threshold')
+    memory_conf = get_memory_usage_params()
+    _memory_max_usage = memory_conf.max_usage
+    _memory_usage_check_interval = memory_conf.check_interval
+    _memory_usage_recovery_threshold = memory_conf.recovery_threshold
 
     worker._memory_max_usage = int(os.environ.get('RC_GUNICORN_MEMORY_MAX_USAGE', '')
                                    or _memory_max_usage)
     worker._memory_usage_check_interval = int(os.environ.get('RC_GUNICORN_MEMORY_USAGE_CHECK_INTERVAL', '')
                                               or _memory_usage_check_interval)
     worker._memory_usage_recovery_threshold = float(os.environ.get('RC_GUNICORN_MEMORY_USAGE_RECOVERY_THRESHOLD', '')
                                                     or _memory_usage_recovery_threshold)
 
     # register memory last check time, with some random offset so we don't recycle all
     # at once
     worker._last_memory_check_time = _time_with_offset(_memory_usage_check_interval)
 
     if _memory_max_usage:
-        server.log.info("[%-10s] WORKER spawned with max memory set at %s", worker.pid,
+        server.log.info("pid=[%-10s] WORKER spawned with max memory set at %s", worker.pid,
                         _format_data_size(_memory_max_usage))
     else:
-        server.log.info("[%-10s] WORKER spawned", worker.pid)
+        server.log.info("pid=[%-10s] WORKER spawned", worker.pid)
 
 
 def pre_exec(server):
     server.log.info("Forked child, re-executing.")
 
 
 def on_starting(server):
     server_lbl = '{} {}'.format(server.proc_name, server.address)
     server.log.info("Server %s is starting.", server_lbl)
 
 
 def when_ready(server):
     server.log.info("Server %s is ready. Spawning workers", server)
 
 
 def on_reload(server):
     pass
 
 
 def _format_data_size(size, unit="B", precision=1, binary=True):
     """Format a number using SI units (kilo, mega, etc.).
 
     ``size``: The number as a float or int.
 
     ``unit``: The unit name in plural form. Examples: "bytes", "B".
 
     ``precision``: How many digits to the right of the decimal point. Default
     is 1.  0 suppresses the decimal point.
 
     ``binary``: If false, use base-10 decimal prefixes (kilo = K = 1000).
     If true, use base-2 binary prefixes (kibi = Ki = 1024).
 
     ``full_name``: If false (default), use the prefix abbreviation ("k" or
     "Ki").  If true, use the full prefix ("kilo" or "kibi"). If false,
     use abbreviation ("k" or "Ki").
 
     """
 
     if not binary:
         base = 1000
         multiples = ('', 'k', 'M', 'G', 'T', 'P', 'E', 'Z', 'Y')
     else:
         base = 1024
         multiples = ('', 'Ki', 'Mi', 'Gi', 'Ti', 'Pi', 'Ei', 'Zi', 'Yi')
 
     sign = ""
     if size > 0:
         m = int(math.log(size, base))
     elif size < 0:
         sign = "-"
         size = -size
         m = int(math.log(size, base))
     else:
         m = 0
     if m > 8:
         m = 8
 
     if m == 0:
         precision = '%.0f'
     else:
         precision = '%%.%df' % precision
 
     size = precision % (size / math.pow(base, m))
 
     return '%s%s %s%s' % (sign, size.strip(), multiples[m], unit)
 
 
 def _check_memory_usage(worker):
-    memory_max_usage = worker._memory_max_usage
-    if not memory_max_usage:
+    _memory_max_usage = worker._memory_max_usage
+    if not _memory_max_usage:
         return
 
-    memory_usage_check_interval = worker._memory_usage_check_interval
-    memory_usage_recovery_threshold = memory_max_usage * worker._memory_usage_recovery_threshold
+    _memory_usage_check_interval = worker._memory_usage_check_interval
+    _memory_usage_recovery_threshold = memory_max_usage * worker._memory_usage_recovery_threshold
 
     elapsed = time.time() - worker._last_memory_check_time
-    if elapsed > memory_usage_check_interval:
+    if elapsed > _memory_usage_check_interval:
         mem_usage = _get_process_rss()
-        if mem_usage and mem_usage > memory_max_usage:
+        if mem_usage and mem_usage > _memory_max_usage:
             worker.log.info(
                 "memory usage %s > %s, forcing gc",
-                _format_data_size(mem_usage), _format_data_size(memory_max_usage))
+                _format_data_size(mem_usage), _format_data_size(_memory_max_usage))
             # Try to clean it up by forcing a full collection.
             gc.collect()
             mem_usage = _get_process_rss()
-            if mem_usage > memory_usage_recovery_threshold:
+            if mem_usage > _memory_usage_recovery_threshold:
                 # Didn't clean up enough, we'll have to terminate.
                 worker.log.warning(
                     "memory usage %s > %s after gc, quitting",
-                    _format_data_size(mem_usage), _format_data_size(memory_max_usage))
+                    _format_data_size(mem_usage), _format_data_size(_memory_max_usage))
                 # This will cause worker to auto-restart itself
                 worker.alive = False
         worker._last_memory_check_time = time.time()
 
 
 def worker_int(worker):
-    worker.log.info("[%-10s] worker received INT or QUIT signal", worker.pid)
+    worker.log.info("pid=[%-10s] worker received INT or QUIT signal", worker.pid)
 
     # get traceback info, on worker crash
-    id2name = dict([(th.ident, th.name) for th in threading.enumerate()])
+    def get_thread_id(t_id):
+        id2name = dict([(th.ident, th.name) for th in threading.enumerate()])
+        return id2name.get(t_id, "unknown_thread_id")
+
     code = []
-    for thread_id, stack in sys._current_frames().items():
+    for thread_id, stack in sys._current_frames().items():  # noqa
         code.append(
-            "\n# Thread: %s(%d)" % (id2name.get(thread_id, ""), thread_id))
+            "\n# Thread: %s(%d)" % (get_thread_id(thread_id), thread_id))
         for fname, lineno, name, line in traceback.extract_stack(stack):
             code.append('File: "%s", line %d, in %s' % (fname, lineno, name))
             if line:
                 code.append("  %s" % (line.strip()))
     worker.log.debug("\n".join(code))
 
 
 def worker_abort(worker):
-    worker.log.info("[%-10s] worker received SIGABRT signal", worker.pid)
+    worker.log.info("pid=[%-10s] worker received SIGABRT signal", worker.pid)
 
 
 def worker_exit(server, worker):
-    worker.log.info("[%-10s] worker exit", worker.pid)
+    worker.log.info("pid=[%-10s] worker exit", worker.pid)
 
 
 def child_exit(server, worker):
-    worker.log.info("[%-10s] worker child exit", worker.pid)
+    worker.log.info("pid=[%-10s] worker child exit", worker.pid)
 
 
 def pre_request(worker, req):
     worker.start_time = time.time()
     worker.log.debug(
         "GNCRN PRE  WORKER [cnt:%s]: %s %s", worker.nr, req.method, req.path)
 
 
 def post_request(worker, req, environ, resp):
     total_time = time.time() - worker.start_time
     # Gunicorn sometimes has problems with reading the status_code
     status_code = getattr(resp, 'status_code', '')
     worker.log.debug(
         "GNCRN POST WORKER [cnt:%s]: %s %s resp: %s, Load Time: %.4fs",
         worker.nr, req.method, req.path, status_code, total_time)
     _check_memory_usage(worker)
 
 
 def _filter_proxy(ip):
     """
     Passed in IP addresses in HEADERS can be in a special format of multiple
     ips. Those comma separated IPs are passed from various proxies in the
     chain of request processing. The left-most being the original client.
     We only care about the first IP which came from the org. client.
 
     :param ip: ip string from headers
     """
     if ',' in ip:
         _ips = ip.split(',')
         _first_ip = _ips[0].strip()
         return _first_ip
     return ip
 
 
 def _filter_port(ip):
     """
     Removes a port from ip, there are 4 main cases to handle here.
     - ipv4 eg. 127.0.0.1
     - ipv6 eg. ::1
     - ipv4+port eg. 127.0.0.1:8080
     - ipv6+port eg. [::1]:8080
 
     :param ip:
     """
     def is_ipv6(ip_addr):
         if hasattr(socket, 'inet_pton'):
             try:
                 socket.inet_pton(socket.AF_INET6, ip_addr)
             except socket.error:
                 return False
         else:
             return False
         return True
 
     if ':' not in ip:  # must be ipv4 pure ip
         return ip
 
     if '[' in ip and ']' in ip:  # ipv6 with port
         return ip.split(']')[0][1:].lower()
 
     # must be ipv6 or ipv4 with port
     if is_ipv6(ip):
         return ip
     else:
         ip, _port = ip.split(':')[:2]  # means ipv4+port
         return ip
 
 
 def get_ip_addr(environ):
     proxy_key = 'HTTP_X_REAL_IP'
     proxy_key2 = 'HTTP_X_FORWARDED_FOR'
     def_key = 'REMOTE_ADDR'
-    _filters = lambda x: _filter_port(_filter_proxy(x))
+
+    def _filters(x):
+        return _filter_port(_filter_proxy(x))
 
     ip = environ.get(proxy_key)
     if ip:
         return _filters(ip)
 
     ip = environ.get(proxy_key2)
     if ip:
         return _filters(ip)
 
     ip = environ.get(def_key, '0.0.0.0')
     return _filters(ip)
 
 
 class RhodeCodeLogger(Logger):
     """
     Custom Logger that allows some customization that gunicorn doesn't allow
     """
 
     datefmt = r"%Y-%m-%d %H:%M:%S"
 
     def __init__(self, cfg):
         Logger.__init__(self, cfg)
 
     def now(self):
         """ return date in RhodeCode Log format """
         now = time.time()
         msecs = int((now - int(now)) * 1000)
         return time.strftime(self.datefmt, time.localtime(now)) + '.{0:03d}'.format(msecs)
 
     def atoms(self, resp, req, environ, request_time):
         """ Gets atoms for log formatting.
         """
         status = resp.status
         if isinstance(status, str):
             status = status.split(None, 1)[0]
         atoms = {
             'h': get_ip_addr(environ),
             'l': '-',
             'u': self._get_user(environ) or '-',
             't': self.now(),
             'r': "%s %s %s" % (environ['REQUEST_METHOD'],
                                environ['RAW_URI'],
                                environ["SERVER_PROTOCOL"]),
             's': status,
             'm': environ.get('REQUEST_METHOD'),
             'U': environ.get('PATH_INFO'),
             'q': environ.get('QUERY_STRING'),
             'H': environ.get('SERVER_PROTOCOL'),
             'b': getattr(resp, 'sent', None) is not None and str(resp.sent) or '-',
             'B': getattr(resp, 'sent', None),
             'f': environ.get('HTTP_REFERER', '-'),
             'a': environ.get('HTTP_USER_AGENT', '-'),
             'T': request_time.seconds,
             'D': (request_time.seconds * 1000000) + request_time.microseconds,
             'M': (request_time.seconds * 1000) + int(request_time.microseconds/1000),
             'L': "%d.%06d" % (request_time.seconds, request_time.microseconds),
             'p': "<%s>" % os.getpid()
         }
 
         # add request headers
         if hasattr(req, 'headers'):
             req_headers = req.headers
         else:
             req_headers = req
 
         if hasattr(req_headers, "items"):
             req_headers = req_headers.items()
 
         atoms.update({"{%s}i" % k.lower(): v for k, v in req_headers})
 
         resp_headers = resp.headers
         if hasattr(resp_headers, "items"):
             resp_headers = resp_headers.items()
 
         # add response headers
         atoms.update({"{%s}o" % k.lower(): v for k, v in resp_headers})
 
         # add environ variables
         environ_variables = environ.items()
         atoms.update({"{%s}e" % k.lower(): v for k, v in environ_variables})
 
         return atoms
 
+
 logger_class = RhodeCodeLogger

configs/production.ini

@@ -1,238 +1,167 b''
 #
 
 ; #################################
 ; RHODECODE VCSSERVER CONFIGURATION
 ; #################################
 
 [server:main]
 ; COMMON HOST/IP CONFIG
 host = 127.0.0.1
-port = 9900
+port = 10010
 
 
 ; ###########################
 ; GUNICORN APPLICATION SERVER
 ; ###########################
 
 ; run with gunicorn --paste rhodecode.ini
 
 ; Module to use, this setting shouldn't be changed
 use = egg:gunicorn#main
 
-; Sets the number of process workers. More workers means more concurrent connections
-; RhodeCode can handle at the same time. Each additional worker also it increases
-; memory usage as each has it's own set of caches.
-; Recommended value is (2 * NUMBER_OF_CPUS + 1), eg 2CPU = 5 workers, but no more
-; than 8-10 unless for really big deployments .e.g 700-1000 users.
-; `instance_id = *` must be set in the [app:main] section below (which is the default)
-; when using more than 1 worker.
-workers = 2
-
-; Gunicorn access log level
-loglevel = info
-
-; Process name visible in process list
-proc_name = rhodecode_vcsserver
-
-; Type of worker class, one of `sync`, `gevent`
-; currently `sync` is the only option allowed.
-worker_class = sync
-
-; The maximum number of simultaneous clients. Valid only for gevent
-worker_connections = 10
-
-; Max number of requests that worker will handle before being gracefully restarted.
-; Prevents memory leaks, jitter adds variability so not all workers are restarted at once.
-max_requests = 1000
-max_requests_jitter = 30
-
-; Amount of time a worker can spend with handling a request before it
-; gets killed and restarted. By default set to 21600 (6hrs)
-; Examples: 1800 (30min), 3600 (1hr), 7200 (2hr), 43200 (12h)
-timeout = 21600
-
-; The maximum size of HTTP request line in bytes.
-; 0 for unlimited
-limit_request_line = 0
-
-; Limit the number of HTTP headers fields in a request.
-; By default this value is 100 and can't be larger than 32768.
-limit_request_fields = 32768
-
-; Limit the allowed size of an HTTP request header field.
-; Value is a positive number or 0.
-; Setting it to 0 will allow unlimited header field sizes.
-limit_request_field_size = 0
-
-; Timeout for graceful workers restart.
-; After receiving a restart signal, workers have this much time to finish
-; serving requests. Workers still alive after the timeout (starting from the
-; receipt of the restart signal) are force killed.
-; Examples: 1800 (30min), 3600 (1hr), 7200 (2hr), 43200 (12h)
-graceful_timeout = 21600
-
-# The number of seconds to wait for requests on a Keep-Alive connection.
-# Generally set in the 1-5 seconds range.
-keepalive = 2
-
-; Maximum memory usage that each worker can use before it will receive a
-; graceful restart signal 0 = memory monitoring is disabled
-; Examples: 268435456 (256MB), 536870912 (512MB)
-; 1073741824 (1GB), 2147483648 (2GB), 4294967296 (4GB)
-memory_max_usage = 0
-
-; How often in seconds to check for memory usage for each gunicorn worker
-memory_usage_check_interval = 60
-
-; Threshold value for which we don't recycle worker if GarbageCollection
-; frees up enough resources. Before each restart we try to run GC on worker
-; in case we get enough free memory after that, restart will not happen.
-memory_usage_recovery_threshold = 0.8
-
-
 [app:main]
 ; The %(here)s variable will be replaced with the absolute path of parent directory
 ; of this file
 ; Each option in the app:main can be override by an environmental variable
 ;
 ;To override an option:
 ;
 ;RC_<KeyName>
 ;Everything should be uppercase, . and - should be replaced by _.
 ;For example, if you have these configuration settings:
 ;rc_cache.repo_object.backend = foo
 ;can be overridden by
 ;export RC_CACHE_REPO_OBJECT_BACKEND=foo
 
 use = egg:rhodecode-vcsserver
 
 ; Pyramid default locales, we need this to be set
 #pyramid.default_locale_name = en
 
 ; default locale used by VCS systems
 #locale = en_US.UTF-8
 
 ; path to binaries for vcsserver, it should be set by the installer
 ; at installation time, e.g /home/user/.rccontrol/vcsserver-1/profile/bin
 ; it can also be a path to nix-build output in case of development
 core.binary_dir = ""
 
 ; Custom exception store path, defaults to TMPDIR
 ; This is used to store exception from RhodeCode in shared directory
 #exception_tracker.store_path =
 
 ; #############
 ; DOGPILE CACHE
 ; #############
 
 ; Default cache dir for caches. Putting this into a ramdisk can boost performance.
 ; eg. /tmpfs/data_ramdisk, however this directory might require large amount of space
 #cache_dir = %(here)s/data
 
 ; ***************************************
 ; `repo_object` cache, default file based
 ; ***************************************
 
 ; `repo_object` cache settings for vcs methods for repositories
 #rc_cache.repo_object.backend = dogpile.cache.rc.file_namespace
 
 ; cache auto-expires after N seconds
 ; Examples: 86400 (1Day), 604800 (7Days), 1209600 (14Days), 2592000 (30days), 7776000 (90Days)
 #rc_cache.repo_object.expiration_time = 2592000
 
 ; file cache store path. Defaults to `cache_dir =` value or tempdir if both values are not set
 #rc_cache.repo_object.arguments.filename = /tmp/vcsserver_cache_repo_object.db
 
 ; ***********************************************************
 ; `repo_object` cache with redis backend
 ; recommended for larger instance, and for better performance
 ; ***********************************************************
 
 ; `repo_object` cache settings for vcs methods for repositories
 #rc_cache.repo_object.backend = dogpile.cache.rc.redis_msgpack
 
 ; cache auto-expires after N seconds
 ; Examples: 86400 (1Day), 604800 (7Days), 1209600 (14Days), 2592000 (30days), 7776000 (90Days)
 #rc_cache.repo_object.expiration_time = 2592000
 
 ; redis_expiration_time needs to be greater then expiration_time
 #rc_cache.repo_object.arguments.redis_expiration_time = 3592000
 
 #rc_cache.repo_object.arguments.host = localhost
 #rc_cache.repo_object.arguments.port = 6379
 #rc_cache.repo_object.arguments.db = 5
 #rc_cache.repo_object.arguments.socket_timeout = 30
 ; more Redis options: https://dogpilecache.sqlalchemy.org/en/latest/api.html#redis-backends
 #rc_cache.repo_object.arguments.distributed_lock = true
 
 ; auto-renew lock to prevent stale locks, slower but safer. Use only if problems happen
 #rc_cache.repo_object.arguments.lock_auto_renewal = true
 
 ; Statsd client config, this is used to send metrics to statsd
 ; We recommend setting statsd_exported and scrape them using Promethues
 #statsd.enabled = false
 #statsd.statsd_host = 0.0.0.0
 #statsd.statsd_port = 8125
 #statsd.statsd_prefix =
 #statsd.statsd_ipv6 = false
 
 ; configure logging automatically at server startup set to false
 ; to use the below custom logging config.
 ; RC_LOGGING_FORMATTER
 ; RC_LOGGING_LEVEL
 ; env variables can control the settings for logging in case of autoconfigure
 
 #logging.autoconfigure = true
 
 ; specify your own custom logging config file to configure logging
 #logging.logging_conf_file = /path/to/custom_logging.ini
 
 ; #####################
 ; LOGGING CONFIGURATION
 ; #####################
 
 [loggers]
 keys = root, vcsserver
 
 [handlers]
 keys = console
 
 [formatters]
 keys = generic, json
 
 ; #######
 ; LOGGERS
 ; #######
 [logger_root]
 level = NOTSET
 handlers = console
 
 [logger_vcsserver]
 level = INFO
 handlers =
 qualname = vcsserver
 propagate = 1
 
 ; ########
 ; HANDLERS
 ; ########
 
 [handler_console]
 class = StreamHandler
 args = (sys.stderr, )
 level = INFO
 ; To enable JSON formatted logs replace 'generic' with 'json'
 ; This allows sending properly formatted logs to grafana loki or elasticsearch
 formatter = generic
 
 ; ##########
 ; FORMATTERS
 ; ##########
 
 [formatter_generic]
 format = %(asctime)s.%(msecs)03d [%(process)d] %(levelname)-5.5s [%(name)s] %(message)s
 datefmt = %Y-%m-%d %H:%M:%S
 
 [formatter_json]
 format = %(timestamp)s %(levelname)s %(name)s %(message)s %(req_id)s
 class = vcsserver.lib._vendor.jsonlogger.JsonFormatter