Bump version to 9.5.5, implement exponential backoff for output client initialization, update http_auth format, and add debug logging for OpenSearch connections

CHANGELOG.md

@@ -1,5 +1,16 @@
 # Changelog
 
+## 9.5.5
+
+### Fixed
+
+- Output client initialization now retries up to 4 times with exponential backoff before exiting. This fixes persistent `Connection refused` errors in Docker when OpenSearch or Elasticsearch is momentarily unavailable at startup.
+- Use tuple format for `http_auth` in OpenSearch and Elasticsearch connections, matching the documented convention and avoiding potential issues if the password contains a colon.
+
+### Changes
+
+- Added debug logging of host and SSL settings when connecting to OpenSearch.
+
 ## 9.5.4
 
 ### Fixed

parsedmarc/cli.py

@@ -9,6 +9,7 @@ import logging
 import os
 import signal
 import sys
+import time
 from argparse import ArgumentParser, Namespace
 from configparser import ConfigParser
 from glob import glob
@@ -1849,15 +1850,30 @@ def _main():
 
     logger.info("Starting parsedmarc")
 
-    # Initialize output clients
-    try:
-        clients = _init_output_clients(opts)
-    except ConfigurationError as e:
-        logger.critical(str(e))
-        exit(1)
-    except Exception as error_:
-        logger.error("Output client error: {0}".format(error_))
-        exit(1)
+    # Initialize output clients (with retry for transient connection errors)
+    max_retries = 4
+    retry_delay = 5
+    for attempt in range(max_retries + 1):
+        try:
+            clients = _init_output_clients(opts)
+            break
+        except ConfigurationError as e:
+            logger.critical(str(e))
+            exit(1)
+        except Exception as error_:
+            if attempt < max_retries:
+                logger.warning(
+                    "Output client error (attempt %d/%d, retrying in %ds): %s",
+                    attempt + 1,
+                    max_retries + 1,
+                    retry_delay,
+                    error_,
+                )
+                time.sleep(retry_delay)
+                retry_delay *= 2
+            else:
+                logger.error("Output client error: {0}".format(error_))
+                exit(1)
 
     file_paths = []
     mbox_paths = []

parsedmarc/constants.py

@@ -1,3 +1,3 @@
-__version__ = "9.5.4"
+__version__ = "9.5.5"
 
 USER_AGENT = f"parsedmarc/{__version__}"

parsedmarc/elastic.py

@@ -299,7 +299,7 @@ def set_hosts(
         else:
             conn_params["verify_certs"] = True
     if username and password:
-        conn_params["http_auth"] = username + ":" + password
+        conn_params["http_auth"] = (username, password)
     if api_key:
         conn_params["api_key"] = api_key
     connections.create_connection(**conn_params)

parsedmarc/opensearch.py

@@ -298,6 +298,7 @@ def set_hosts(
     """
     if not isinstance(hosts, list):
         hosts = [hosts]
+    logger.debug("Connecting to OpenSearch: hosts=%s, use_ssl=%s", hosts, use_ssl)
     conn_params = {"hosts": hosts, "timeout": timeout}
     if use_ssl:
         conn_params["use_ssl"] = True
@@ -323,7 +324,7 @@ def set_hosts(
         conn_params["connection_class"] = RequestsHttpConnection
     elif normalized_auth_type == "basic":
         if username and password:
-            conn_params["http_auth"] = username + ":" + password
+            conn_params["http_auth"] = (username, password)
         if api_key:
             conn_params["api_key"] = api_key
     else: