argus-alm 0.15.3__tar.gz → 0.15.5__tar.gz

{argus_alm-0.15.3 → argus_alm-0.15.5}/.gitignore

@@ -137,3 +137,4 @@ argus_web.yaml
 .vscode
 /profile/*
 !/profile/.gitkeep
+argus/_version.py

{argus_alm-0.15.3 → argus_alm-0.15.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: argus-alm
-Version: 0.15.3
+Version: 0.15.5
 Summary: Argus
 Author-email: Alexey Kartashov <alexey.kartashov@scylladb.com>, Łukasz Sójka <lukasz.sojka@scylladb.com>
 License-Expression: Apache-2.0
@@ -13,16 +13,16 @@ Requires-Dist: requests>=2.26.0
 Requires-Dist: click>=8.1.3
 Provides-Extra: web-backend
 Requires-Dist: PyYAML~=6.0.0; extra == "web-backend"
-Requires-Dist: scylla-driver~=3.26.8; extra == "web-backend"
+Requires-Dist: scylla-driver>=3.26.8; extra == "web-backend"
 Requires-Dist: Flask~=3.0.0; extra == "web-backend"
 Requires-Dist: Flask-WTF~=1.0.0; extra == "web-backend"
 Requires-Dist: Flask-Login~=0.5.0; extra == "web-backend"
 Requires-Dist: humanize~=3.13.1; extra == "web-backend"
 Requires-Dist: python-magic~=0.4.24; extra == "web-backend"
 Requires-Dist: uwsgi~=2.0.20; extra == "web-backend"
-Requires-Dist: python-jenkins~=1.7.0; extra == "web-backend"
+Requires-Dist: python-jenkins>=1.7.0; extra == "web-backend"
 Requires-Dist: python-slugify~=6.1.1; extra == "web-backend"
-Requires-Dist: pygithub~=2.6.1; extra == "web-backend"
+Requires-Dist: pygithub>=2.6.1; extra == "web-backend"
 Requires-Dist: boto3~=1.38.9; extra == "web-backend"
 Provides-Extra: docker-image
 Requires-Dist: supervisor~=4.2.4; extra == "docker-image"

{argus_alm-0.15.3 → argus_alm-0.15.5}/argus/_version.py

@@ -17,5 +17,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '0.15.3'
-__version_tuple__ = version_tuple = (0, 15, 3)
+__version__ = version = '0.15.5'
+__version_tuple__ = version_tuple = (0, 15, 5)

{argus_alm-0.15.3 → argus_alm-0.15.5}/argus/client/base.py

@@ -96,14 +96,17 @@ class ArgusClient:
         }
 
     def get(self, endpoint: str, location_params: dict[str, str] = None, params: dict = None) -> requests.Response:
+        url = self.get_url_for_endpoint(
+            endpoint=endpoint,
+            location_params=location_params
+        )
+        LOGGER.debug("GET Request: %s, params: %s", url, params)
         response = requests.get(
-            url=self.get_url_for_endpoint(
-                endpoint=endpoint,
-                location_params=location_params
-            ),
+            url=url,
             params=params,
             headers=self.request_headers
         )
+        LOGGER.debug("GET Response: %s %s", response.status_code, response.url)
 
         return response
 
@@ -114,15 +117,18 @@ class ArgusClient:
         params: dict = None,
         body: dict = None,
     ) -> requests.Response:
+        url = self.get_url_for_endpoint(
+            endpoint=endpoint,
+            location_params=location_params
+        )
+        LOGGER.debug("POST Request: %s, params: %s, body: %s", url, params, body)
         response = requests.post(
-            url=self.get_url_for_endpoint(
-                endpoint=endpoint,
-                location_params=location_params
-            ),
+            url=url,
             params=params,
             json=body,
             headers=self.request_headers,
         )
+        LOGGER.debug("POST Response: %s %s", response.status_code, response.url)
 
         return response
 

{argus_alm-0.15.3 → argus_alm-0.15.5}/argus/client/generic/cli.py

@@ -32,13 +32,14 @@ def cli():
 @click.option("--build-id", required=True, help="Unique job identifier in the build system, e.g. scylla-master/group/job for jenkins (The full path)")
 @click.option("--build-url", required=True, help="Job URL in the build system")
 @click.option("--started-by", required=True, help="Username of the user who started the job")
+@click.option("--sub-type", required=False, help="Sub-type of the generic test: pytest, dtest")
 @click.option("--scylla-version", required=False, default=None, help="Version of Scylla used for this job")
 @click.option("--extra-headers", default={}, type=click.UNPROCESSED, callback=validate_extra_headers, help="extra headers to pass to argus, should be in json format", envvar='ARGUS_EXTRA_HEADERS')
-def submit_run(api_key: str, base_url: str, id: str, build_id: str, build_url: str, started_by: str, scylla_version: str = None, extra_headers: dict | None = None):
+def submit_run(api_key: str, base_url: str, id: str, build_id: str, build_url: str, started_by: str, sub_type: str = None, scylla_version: str = None, extra_headers: dict | None = None):
     LOGGER.info("Submitting %s (%s) to Argus...", build_id, id)
     client = ArgusGenericClient(auth_token=api_key, base_url=base_url, extra_headers=extra_headers)
     client.submit_generic_run(build_id=build_id, run_id=id, started_by=started_by,
-                              build_url=build_url, scylla_version=scylla_version)
+                              build_url=build_url, scylla_version=scylla_version, sub_type=sub_type)
     LOGGER.info("Done.")
 
 

{argus_alm-0.15.3 → argus_alm-0.15.5}/argus/client/generic/client.py

@@ -15,12 +15,13 @@ class ArgusGenericClient(ArgusClient):
     def __init__(self, auth_token: str, base_url: str, api_version="v1", extra_headers: dict | None = None) -> None:
         super().__init__(auth_token, base_url, api_version, extra_headers=extra_headers)
 
-    def submit_generic_run(self, build_id: str, run_id: str, started_by: str, build_url: str, scylla_version: str | None = None):
+    def submit_generic_run(self, build_id: str, run_id: str, started_by: str, build_url: str, sub_type: str = None, scylla_version: str | None = None):
         request_body = {
             "build_id": build_id,
             "run_id": run_id,
             "started_by": started_by,
             "build_url": build_url,
+            "sub_type": sub_type,
             "scylla_version": scylla_version
         }
 

{argus_alm-0.15.3 → argus_alm-0.15.5}/argus/client/sct/client.py

@@ -152,7 +152,7 @@ class ArgusSCTClient(ArgusClient):
         )
         self.check_response(response)
 
-    def create_resource(self, name: str, resource_type: str, public_ip: str, private_ip: str,
+    def create_resource(self, name: str, resource_type: str, public_ip: str, private_ip: str, instance_type: str,
                         region: str, provider: str, dc_name: str, rack_name: str, shards_amount: int, state=ResourceState.RUNNING) -> None:
         """
             Creates a cloud resource record in argus.
@@ -167,6 +167,7 @@ class ArgusSCTClient(ArgusClient):
                     "state": state,
                     "resource_type": resource_type,
                     "instance_details": {
+                        "instance_type": instance_type,
                         "provider": provider,
                         "region": region,
                         "dc_name": dc_name,

{argus_alm-0.15.3 → argus_alm-0.15.5}/argus/common/enums.py

@@ -22,6 +22,22 @@ class TestStatus(str, Enum):
     NOT_RUN = "not_run"
 
 
+class PytestStatus(str, Enum):
+    ERROR = "error"
+    PASSED = "passed"
+    FAILURE = "failure"
+    SKIPPED = "skipped"
+    XFAILED = "xfailed"
+    XPASS = "xpass"
+    # Following statuses reflect test & teardown state
+    # Example: passed & error means the test passed
+    # but the tearDown stage errored
+    PASSED_ERROR = "passed & error"
+    FAILURE_ERROR = "failure & error"
+    SKIPPED_ERROR = "skipped & error"
+    ERROR_ERROR = "error & error"
+
+
 class TestInvestigationStatus(str, Enum):
     NOT_INVESTIGATED = "not_investigated"
     IN_PROGRESS = "in_progress"

{argus_alm-0.15.3 → argus_alm-0.15.5}/argusAI/deployment/argusai_event_similarity_processor.service

@@ -6,7 +6,7 @@ After=network.target
 Type=simple
 User=argus
 Group=argus
-WorkingDirectory=/home/argus/app
+WorkingDirectory=/home/argus/argusAI
 Environment="PYTHONPATH=."
 Environment="DAYS_BACK=120"
 ExecStart=/usr/bin/uv --project argusAI/pyproject.toml run argusAI/event_similarity_processor.py

{argus_alm-0.15.3 → argus_alm-0.15.5}/argusAI/event_similarity_processor.py

@@ -1,6 +1,8 @@
 import os
 from collections import namedtuple
 from pathlib import Path
+
+import cassandra
 import chromadb
 from threading import Thread, Event
 import time
@@ -9,7 +11,7 @@ import datetime
 from typing import Dict, Set, List, Tuple, Union
 import uuid
 
-from chromadb.api.types import IncludeEnum, Embeddings
+from chromadb.api.types import Embeddings
 from chromadb.utils.embedding_functions.onnx_mini_lm_l6_v2 import ONNXMiniLM_L6_V2
 from chromadb.config import Settings
 
@@ -24,6 +26,7 @@ SLEEP_INTERVAL: int = 300  # 5 minutes
 SIMILARITY_THRESHOLD: float = 0.90
 MAX_SIMILARS: int = 200
 DAYS_BACK: int = int(os.getenv("DAYS_BACK", "120"))
+RECREATE_EMBEDDINGS: bool = os.getenv("RECREATE_EMBEDDINGS", "false").lower() in ("true", "1", "yes")
 
 logging.basicConfig(
     level=logging.INFO,
@@ -98,13 +101,22 @@ class EventSimilarityProcessor:
 
         LOGGER.info(f"Finding similarities for run_id: {run_id}, event_index: {event_index}, type: {event_type}")
 
-        similar: Dict = collection.query(
+        similar = collection.query(
             query_embeddings=[embedding],
             n_results=max_similars,
             where={"run_id": {"$ne": str(run_id)}},
             include=[
-                IncludeEnum.metadatas,
-                IncludeEnum.distances,
+                "metadatas",
+                "distances",
+            ],
+        )
+        duplicate = collection.query(
+            query_embeddings=[embedding],
+            n_results=100,
+            where={"run_id": str(run_id)},
+            include=[
+                "metadatas",
+                "distances",
             ],
         )
 
@@ -117,17 +129,24 @@ class EventSimilarityProcessor:
                 if len(similar_map) >= max_similars:
                     break
 
+        duplicate_list: List[int] = []
+        for j, distance in enumerate(duplicate["distances"][0]):
+            if 1 - distance >= 0.95 and event_index != duplicate["metadatas"][0][j]["idx"]:
+                duplicate_list.append(duplicate["metadatas"][0][j]["idx"])
+
         table = ErrorEventEmbeddings if event_type == EventType.ERROR else CriticalEventEmbeddings
         try:
             self._db.execute(
-                f"UPDATE {table.__table_name__} SET similars_map = ? WHERE run_id = ? AND event_index = ?",
-                (similar_map, run_id, event_index),
+                f"UPDATE {table.__table_name__} SET similars_map = ?, duplicates_list = ? WHERE run_id = ? AND event_index = ?",
+                (similar_map, duplicate_list, run_id, event_index),
             )
             LOGGER.info(
                 f"Found {len(similar_map)} similar {event_type} events for run_id: {run_id}, event_index: {event_index}"
             )
         except Exception as e:  # noqa: BLE001
-            LOGGER.error(f"Failed to update similars_map for run_id: {run_id}, event_index: {event_index}: {e}")
+            LOGGER.error(
+                f"Failed to update similars_map/duplicates_list for run_id: {run_id}, event_index: {event_index}: {e}"
+            )
 
     def load_events(
         self, event_type: str, cutoff_time: datetime.datetime, batch_size: int = 5000
@@ -148,7 +167,9 @@ class EventSimilarityProcessor:
         collection: chromadb.Collection = self._collections[event_type]
         table = ErrorEventEmbeddings if event_type == EventType.ERROR else CriticalEventEmbeddings
         similars_to_update: List[EmbeddingEntry] = []
-
+        if RECREATE_EMBEDDINGS:
+            LOGGER.info(f"Skipping loading of {event_type} events due to RECREATE_EMBEDDINGS setting")
+            return processed_runs, similars_to_update
         LOGGER.info(f"Loading processed {event_type} events...")
         query: str = f"SELECT * FROM {table.__table_name__} WHERE start_time > ? ALLOW FILTERING BYPASS CACHE"
         rows = self._db.execute(query, (cutoff_time,))
@@ -168,7 +189,7 @@ class EventSimilarityProcessor:
                 ids.append(f"{emb.run_id}_{emb.event_index}")
                 embeddings_list.append(emb.embedding)
                 metadatas.append({"run_id": str(emb.run_id), "idx": emb.event_index})
-                if emb.similars_map is None:
+                if emb.similars_map is None or emb.duplicates_list is None:
                     similars_to_update.append(
                         EmbeddingEntry(emb.run_id, emb.event_index, emb.embedding, emb.start_time)
                     )
@@ -195,13 +216,19 @@ class EventSimilarityProcessor:
         new_embeddings: Dict[str, List[EmbeddingEntry]] = similars_to_find
         while not stop_event.is_set():
             LOGGER.info(f"Starting event processing cycle; fetching tests since: {cutoff_time}")
-            tests_to_process_result = self._db.execute(
-                f"SELECT id, status FROM {SCTTestRun.table_name()} WHERE start_time > ? ALLOW FILTERING BYPASS CACHE",
-                (cutoff_time,),
-            )
-            tests_to_process: Set[uuid.UUID] = {
-                run.id for run in tests_to_process_result if run.status != TestStatus.PASSED.value
-            } - processed_runs
+            try:
+                tests_to_process_result = self._db.execute(
+                    f"SELECT id, status FROM {SCTTestRun.table_name()} WHERE start_time > ? ALLOW FILTERING BYPASS CACHE",
+                    (cutoff_time,),
+                )
+                tests_to_process: Set[uuid.UUID] = {
+                    run.id for run in tests_to_process_result if run.status != TestStatus.PASSED.value
+                } - processed_runs
+            except cassandra.ReadTimeout:
+                LOGGER.error("Read Timeout while trying to get new events to process - skipping until next iteration")
+                stop_event.wait(timeout=SLEEP_INTERVAL)
+                continue
+
             LOGGER.info(f"tests to process (including in progress) count: {len(tests_to_process)}")
             try:
                 stats: Dict[str, int] = {

{argus_alm-0.15.3 → argus_alm-0.15.5}/argusAI/pyproject.toml

@@ -6,5 +6,6 @@ readme = "README.md"
 requires-python = ">=3.12"
 dependencies = [
     "chromadb>=0.6.3",
+    "flask>=3.1.1",
     "scylla-driver>=3.28.0",
 ]

{argus_alm-0.15.3 → argus_alm-0.15.5}/argusAI/utils/event_message_sanitizer.py

@@ -49,6 +49,7 @@ class MessageSanitizer:
             r"Machine ID|Hostname|Storage|Size on Disk|Message|Disk Size):.*?(?=\n|$)",
             re.MULTILINE,
         )
+        self.report_at_pattern: Pattern = re.compile(r"Please report: at .*\n")
         self.lib_address_pattern: Pattern = re.compile(r"\([^()]+ \+ 0x[0-9a-fA-F]+\)")
         self.module_pattern: Pattern = re.compile(r"Module [^\n]+ from rpm [^\n]+")
         self.stack_header_pattern: Pattern = re.compile(r"Stack trace of thread \d+:")
@@ -62,6 +63,17 @@ class MessageSanitizer:
         self.eol_pattern = re.compile(r"\\n")
         self.backslashes_pattern = re.compile(r"\\")
         self.repetitions_pattern = re.compile(r"(\b\w+\b)(?:\s+\1)+")
+        self.backtrace_unwanted_lines = {
+            "seastar::backtrace",
+            "seastar::print_with_backtrace",
+            "seastar::install_oneshot_signal_handler",
+            "seastar::current_backtrace",
+            "seastar::current_tasktrace",
+            "seastar::memory::cpu_pages::warn_large_allocation",
+            "seastar::memory::allocate_slowpath",
+            "at main.cc:",
+            " ??:",
+        }
 
         self.sanitizers: List[Callable[[str], str]] = [
             self.remove_preface,
@@ -72,8 +84,10 @@ class MessageSanitizer:
             self.remove_ip_addresses,
             self.remove_urls,
             self.remove_metadata,
+            self.remove_backtrace_unwanted_lines,
             self.remove_modules,
             self.remove_message_prefix,
+            self.remove_report_at,
             self.remove_lib_address,
             self.remove_process_ids,
             self.remove_memory_addresses,
@@ -213,3 +227,22 @@ class MessageSanitizer:
 
     def remove_repetitions(self, text: str) -> str:
         return self.repetitions_pattern.sub(r"\1", text)
+
+    def remove_report_at(self, text: str) -> str:
+        return self.report_at_pattern.sub("", text)
+
+    def _remove_inlined_starting(self, text: str) -> str:
+        """Remove inlined starting text that is not useful."""
+        if not text.startswith("(inlined by) "):
+            return text
+        return text.split("(inlined by) ", 1)[-1].strip()
+
+    def remove_backtrace_unwanted_lines(self, text: str) -> str:
+        """Specific lines in backtrace that are not useful and can be removed."""
+        if not "backtrace" in text[:1000]:
+            return text
+        return "\n".join(
+            self._remove_inlined_starting(line)
+            for line in text.splitlines()
+            if not any(problem in line for problem in self.backtrace_unwanted_lines)
+        )