Merge branch 'master' into more-docs-improvements

Author: aantn

src/robusta/core/exceptions.py

@@ -14,3 +14,20 @@ class NoAlertManagerUrlFound(Exception):
     """Exception, when AlertManager url is incorrect"""
 
     pass
+
+
+class SupabaseDnsException(Exception):
+    """Exception for Supabase DNS/connectivity issues (host resolution / firewall).
+
+    Raised when the runner cannot resolve or connect to the configured Supabase URL.
+    """
+    def __init__(self, original_exception: Exception, url: str):
+        message = (
+            f"\n{original_exception.__class__.__name__}: {original_exception}\n"
+            f"Error connecting to {url}\n"
+            f"This is often due to DNS issues or firewall policies - to troubleshoot run in your cluster:\n"
+            f"curl -I {url}"
+        )
+        super().__init__(message)
+        self.original_exception = original_exception
+        self.url = url

src/robusta/core/sinks/robusta/dal/supabase_dal.py

@@ -6,7 +6,6 @@
 import threading
 from typing import Any, Dict, List, Optional, Tuple
 from uuid import uuid4
-
 from cachetools import TTLCache
 import requests
 from postgrest._sync.request_builder import SyncQueryRequestBuilder
@@ -18,6 +17,7 @@
 from supabase.lib.client_options import ClientOptions
 
 from robusta.core.model.cluster_status import ClusterStatus
+from robusta.core.exceptions import SupabaseDnsException
 from robusta.core.model.env_vars import SUPABASE_TIMEOUT_SECONDS
 from robusta.core.model.helm_release import HelmRelease
 from robusta.core.model.jobs import JobInfo
@@ -555,10 +555,24 @@ def publish_helm_releases(self, helm_releases: List[HelmRelease]):
 
     def sign_in(self) -> str:
         logging.info("Supabase dal login")
-        res = self.client.auth.sign_in_with_password({"email": self.email, "password": self.password})
-        self.client.auth.set_session(res.session.access_token, res.session.refresh_token)
-        self.client.postgrest.auth(res.session.access_token)
-        return res.user.id
+        try:
+            res = self.client.auth.sign_in_with_password({"email": self.email, "password": self.password})
+            self.client.auth.set_session(res.session.access_token, res.session.refresh_token)
+            self.client.postgrest.auth(res.session.access_token)
+            return res.user.id
+        except Exception as e:
+            # Check if this is a DNS-related error
+            error_msg = str(e).lower()
+            if any(dns_indicator in error_msg for dns_indicator in [
+                "temporary failure in name resolution",
+                "name resolution",
+                "dns",
+                "name or service not known",
+                "nodename nor servname provided"
+            ]):
+                dns_exc = SupabaseDnsException(e, self.url)
+                raise dns_exc from e
+            raise
 
     def to_db_cluster_status(self, data: ClusterStatus) -> Dict[str, Any]:
         db_cluster_status = data.dict()

src/robusta/model/config.py

@@ -11,6 +11,7 @@
 from robusta.core.pubsub.events_pubsub import EventsPubSub
 from robusta.core.sinks.robusta.robusta_sink import RobustaSink
 from robusta.core.sinks.robusta.robusta_sink_params import RobustaSinkConfigWrapper, RobustaSinkParams
+from robusta.core.exceptions import SupabaseDnsException
 from robusta.core.sinks.sink_base import SinkBase
 from robusta.core.sinks.sink_config import SinkConfigBase
 from robusta.core.sinks.sink_factory import SinkFactory
@@ -91,7 +92,9 @@ def construct_new_sinks(
 
             except Exception as e:
                 has_sink_errors = True
-                logging.error(f"Failed to initialize sink {sink_name}: {e}", exc_info=True)
+                # Don't show trace for Robusta SaaS DNS-related exception to avoid hiding the error in noisy logs
+                exec_info = not isinstance(e, SupabaseDnsException)
+                logging.error(f"Failed to initialize sink {sink_name}: {e}", exc_info=exec_info)
                 if not continue_on_sink_errors:
                     raise
                 # Skip this sink if continue_on_sink_errors is True

src/robusta/runner/config_loader.py

@@ -43,6 +43,7 @@
 from robusta.model.playbook_definition import PlaybookDefinition
 from robusta.utils.cluster_provider_discovery import cluster_provider
 from robusta.utils.file_system_watcher import FileSystemWatcher
+from robusta.core.exceptions import SupabaseDnsException
 
 
 class ConfigLoader:
@@ -271,10 +272,12 @@ def __reload_playbook_packages(self, change_name):
                     str(runner_config.global_config.get("cluster_name", "no_cluster")).encode("utf-8")
                 ).hexdigest()
 
-            except Exception:
+            except Exception as e:
+                print_exception = not isinstance(e, SupabaseDnsException) # Don't show trace for Robusta SaaS DNS-related exception to avoid hiding the error in noisy logs
+                
                 logging.error(
                     "Error (re)loading playbooks/related resources, exiting.",
-                    exc_info=True,
+                    exc_info=print_exception,
                 )
                 # Kill the whole process group (which means this process and all of its descendant
                 # processes). The rest of the runner shutdown happens in robusta.runner.process_setup.