chore: remove NewRelic integration and cleanup related files

docs/infrastructure/troubleshooting.md

@@ -8,27 +8,6 @@ For less urgent issues or general support, you can file a bug with [cloudOps](ht
 
 ## Monitoring & metrics
 
-- New Relic
-  - Overview:
-    [prototype](https://rpm.newrelic.com/accounts/677903/applications/7385291) |
-    [stage](https://rpm.newrelic.com/accounts/677903/applications/14179733) |
-    [prod](https://rpm.newrelic.com/accounts/677903/applications/14179757)
-  - Error analytics:
-    [prototype](https://rpm.newrelic.com/accounts/677903/applications/7385291/filterable_errors) |
-    [stage](https://rpm.newrelic.com/accounts/677903/applications/14179733/filterable_errors) |
-    [prod](https://rpm.newrelic.com/accounts/677903/applications/14179757/filterable_errors)
-  - Web transactions:
-    [prototype](https://rpm.newrelic.com/accounts/677903/applications/7385291/transactions?type=app) |
-    [stage](https://rpm.newrelic.com/accounts/677903/applications/14179733/transactions?type=app) |
-    [prod](https://rpm.newrelic.com/accounts/677903/applications/14179757/transactions?type=app)
-  - Non-web transactions (background tasks):
-    [prototype](https://rpm.newrelic.com/accounts/677903/applications/7385291/transactions?type=other&show_browser=false) |
-    [stage](https://rpm.newrelic.com/accounts/677903/applications/14179733/transactions?type=other&show_browser=false) |
-    [prod](https://rpm.newrelic.com/accounts/677903/applications/14179757/transactions?type=other&show_browser=false)
-  - Postgres/Redis client request stats:
-    [prototype](https://rpm.newrelic.com/accounts/677903/applications/7385291/datastores) |
-    [stage](https://rpm.newrelic.com/accounts/677903/applications/14179733/datastores) |
-    [prod](https://rpm.newrelic.com/accounts/677903/applications/14179757/datastores)
 - Google Cloud Console
   - [prod](https://console.cloud.google.com/kubernetes/list?project=moz-fx-treeherder-prod-c739)
   - [all other deployments](https://console.cloud.google.com/kubernetes/list?project=moz-fx-treeherde-nonprod-34ec)

treeherder/config/settings.py

@@ -26,7 +26,6 @@
 LOGGING_LEVEL = env("LOGGING_LEVEL", default="INFO")
 
 NEW_RELIC_INSIGHTS_API_KEY = env("NEW_RELIC_INSIGHTS_API_KEY", default=None)
-NEW_RELIC_INSIGHTS_API_URL = "https://insights-api.newrelic.com/v1/accounts/677903/query"
 
 # Make this unique, and don't share it with anybody.
 SECRET_KEY = env(

treeherder/etl/classification_loader.py

@@ -2,7 +2,6 @@
 import re
 
 import environ
-import newrelic.agent
 
 from treeherder.model.models import (
     BugJobMap,
@@ -106,15 +105,13 @@ def get_push(self, task_route):
             raise
 
         try:
-            newrelic.agent.add_custom_attribute("project", project)
 
             repository = Repository.objects.get(name=project)
         except Repository.DoesNotExist:
             logger.info("Job with unsupported project: %s", project)
             raise
 
         try:
-            newrelic.agent.add_custom_attribute("revision", revision)
 
             revision_field = "revision__startswith" if len(revision) < 40 else "revision"
             filter_kwargs = {"repository": repository, revision_field: revision}

treeherder/etl/job_loader.py

@@ -2,7 +2,6 @@
 import uuid
 
 import jsonschema
-import newrelic.agent
 import slugid
 from django.conf import settings
 
@@ -67,7 +66,6 @@ def process_job(self, pulse_job, root_url):
             try:
                 with settings.STATSD_CLIENT.timer("process_job_transform"):
                     project = pulse_job["origin"]["project"]
-                    newrelic.agent.add_custom_attribute("project", project)
 
                     repository = Repository.objects.get(name=project)
                     if repository.active_status != "active":
@@ -103,11 +101,6 @@ def validate_revision(self, repository, pulse_job):
         revision_field = "revision__startswith" if len(revision) < 40 else "revision"
         filter_kwargs = {"repository": repository, revision_field: revision}
 
-        if revision_field == "revision__startswith":
-            newrelic.agent.record_custom_event(
-                "short_revision_job_loader",
-                {"error": "Revision <40 chars", "revision": revision, "job": pulse_job},
-            )
 
         if not Push.objects.filter(**filter_kwargs).exists():
             (real_task_id, _) = task_and_retry_ids(pulse_job["taskId"])

treeherder/etl/jobs.py

@@ -5,7 +5,6 @@
 from datetime import datetime
 from hashlib import sha1
 
-import newrelic.agent
 from django.conf import settings
 from django.core.exceptions import ObjectDoesNotExist
 from django.db.utils import IntegrityError
@@ -500,7 +499,6 @@ def store_job_data(repository, original_data):
             # make more fields visible in new relic for the job
             # where we encountered the error
             datum.update(datum.get("job", {}))
-            newrelic.agent.notice_error(attributes=datum)
 
             # skip any jobs that hit errors in these stages.
             continue

treeherder/etl/push_loader.py

@@ -1,7 +1,6 @@
 import logging
 
 import environ
-import newrelic.agent
 from django.core.exceptions import ObjectDoesNotExist
 
 from treeherder.etl.common import to_timestamp
@@ -20,15 +19,12 @@ class PushLoader:
     def process(self, message_body, exchange, root_url):
         transformer = self.get_transformer_class(exchange)(message_body)
         try:
-            newrelic.agent.add_custom_attribute("url", transformer.repo_url)
-            newrelic.agent.add_custom_attribute("branch", transformer.branch)
             repos = Repository.objects
             if transformer.branch:
                 repos = repos.filter(branch__regex=f"(^|,){transformer.branch}($|,)")
             else:
                 repos = repos.filter(branch=None)
             repo = repos.get(url=transformer.repo_url, active_status="active")
-            newrelic.agent.add_custom_attribute("repository", repo.name)
         except ObjectDoesNotExist:
             repo_info = transformer.get_info()
             repo_info.update(
@@ -37,7 +33,6 @@ def process(self, message_body, exchange, root_url):
                     "branch": transformer.branch,
                 }
             )
-            newrelic.agent.record_custom_event("skip_unknown_repository", repo_info)
             logger.warning(
                 "Skipping unsupported repo: %s %s", transformer.repo_url, transformer.branch
             )
@@ -257,7 +252,6 @@ def transform(self, repository):
         return self.fetch_push(url, repository)
 
     def fetch_push(self, url, repository, sha=None):
-        newrelic.agent.add_custom_attribute("sha", sha)
 
         logger.debug("fetching for %s %s", repository, url)
         # there will only ever be one, with this url

treeherder/etl/pushlog.py

@@ -1,7 +1,6 @@
 import logging
 import traceback
 
-import newrelic.agent
 import requests
 from django.core.cache import cache
 
@@ -123,7 +122,6 @@ def run(self, source_url, repository_name, changeset=None, last_push_id=None):
             try:
                 store_push(repository, self.transform_push(push))
             except Exception:
-                newrelic.agent.notice_error()
                 errors.append(
                     {
                         "project": repository,

treeherder/etl/tasks/pulse_tasks.py

@@ -4,7 +4,6 @@
 
 import asyncio
 
-import newrelic.agent
 from django.conf import settings
 
 from treeherder.etl.classification_loader import ClassificationLoader
@@ -25,8 +24,7 @@ def store_pulse_tasks(
     Fetches tasks from Taskcluster
     """
     loop = asyncio.get_event_loop()
-    newrelic.agent.add_custom_attribute("exchange", exchange)
-    newrelic.agent.add_custom_attribute("routing_key", routing_key)
+
     # handle_message expects messages in this format
     with settings.STATSD_CLIENT.timer("pulse_handle_message"):
         runs = loop.run_until_complete(
@@ -50,8 +48,7 @@ def store_pulse_pushes(
     """
     Fetches the pushes pending from pulse exchanges and loads them.
     """
-    newrelic.agent.add_custom_attribute("exchange", exchange)
-    newrelic.agent.add_custom_attribute("routing_key", routing_key)
+
 
     PushLoader().process(body, exchange, root_url)
 
@@ -66,7 +63,5 @@ def store_pulse_tasks_classification(
     By default, it should listen to the Community cluster as classifications
     are only running there for the moment
     """
-    newrelic.agent.add_custom_attribute("exchange", exchange)
-    newrelic.agent.add_custom_attribute("routing_key", routing_key)
 
     ClassificationLoader().process(pulse_job, root_url)

treeherder/etl/tasks/pushlog_tasks.py

@@ -1,4 +1,3 @@
-import newrelic.agent
 from celery import shared_task
 
 from treeherder.etl.pushlog import HgPushlogProcess
@@ -19,6 +18,6 @@ def fetch_hg_push_log(repo_name, repo_url):
     """
     Run a HgPushlog etl process
     """
-    newrelic.agent.add_custom_attribute("repo_name", repo_name)
+
     process = HgPushlogProcess()
     process.run(repo_url + "/json-pushes/?full=1&version=2", repo_name)

treeherder/log_parser/artifactbuildercollection.py

@@ -1,7 +1,5 @@
 import logging
 
-import newrelic.agent
-
 from treeherder.utils.http import make_request
 
 from .artifactbuilders import LogViewerArtifactBuilder, PerformanceDataArtifactBuilder
@@ -86,10 +84,6 @@ def parse(self):
             download_size_in_bytes = int(response.headers.get("Content-Length", -1))
 
             # Temporary annotation of log size to help set thresholds in bug 1295997.
-            newrelic.agent.add_custom_attribute("unstructured_log_size", download_size_in_bytes)
-            newrelic.agent.add_custom_attribute(
-                "unstructured_log_encoding", response.headers.get("Content-Encoding", "None")
-            )
 
             if download_size_in_bytes > MAX_DOWNLOAD_SIZE_IN_BYTES:
                 raise LogSizeError(f"Download size of {download_size_in_bytes} bytes exceeds limit")

treeherder/log_parser/failureline.py

@@ -3,7 +3,6 @@
 from collections import defaultdict
 from itertools import islice
 
-import newrelic.agent
 from django.conf import settings
 from django.db import transaction
 from django.db.utils import DataError, IntegrityError, OperationalError
@@ -119,17 +118,8 @@ def create_group_result(job_log, line):
     # Log to New Relic if it's not in a form we like.  We can enter
     # Bugs to upstream to remedy them.
     if "\\" in group_path or len(group_path) > 255:
-        newrelic.agent.record_custom_event(
-            "malformed_test_group",
-            {
-                "message": "Group paths must be relative, with no backslashes and <255 chars",
-                "group": line["group"],
-                "group_path": group_path,
-                "length": len(group_path),
-                "repository": job_log.job.repository,
-                "job_guid": job_log.job.guid,
-            },
-        )
+        pass
+
     else:
         group, _ = Group.objects.get_or_create(name=group_path[:255])
         duration = line.get("duration", 0)
@@ -157,7 +147,6 @@ def create(job_log, log_list):
     for line in log_list:
         action = line["action"]
         if action not in FailureLine.ACTION_LIST:
-            newrelic.agent.record_custom_event("unsupported_failure_line_action", line)
             # Unfortunately, these errors flood the logs, but we want to report any
             # others that we didn't expect.  We know about the following action we choose
             # to ignore.

treeherder/log_parser/tasks.py

@@ -1,6 +1,5 @@
 import logging
 
-import newrelic.agent
 import simplejson as json
 from celery.exceptions import SoftTimeLimitExceeded
 from requests.exceptions import HTTPError
@@ -20,7 +19,6 @@
 
 @retryable_task(name="log-parser", max_retries=10)
 def parse_logs(job_id, job_log_ids, priority):
-    newrelic.agent.add_custom_attribute("job_id", str(job_id))
 
     job = Job.objects.get(id=job_id)
     job_logs = JobLog.objects.filter(id__in=job_log_ids, job=job)
@@ -41,7 +39,6 @@ def parse_logs(job_id, job_log_ids, priority):
     first_exception = None
     completed_names = set()
     for job_log in job_logs:
-        newrelic.agent.add_custom_attribute(f"job_log_{job_log.name}_url", job_log.url)
         logger.info("parser_task for %s", job_log.id)
 
         # Only parse logs which haven't yet been processed or else failed on the last attempt.
@@ -60,16 +57,13 @@ def parse_logs(job_id, job_log_ids, priority):
             parser(job_log)
         except Exception as e:
             if isinstance(e, SoftTimeLimitExceeded):
-                # stop parsing further logs but raise so NewRelic and
+                # stop parsing further logs but raise so
                 # Papertrail will still show output
                 raise
 
             if first_exception is None:
                 first_exception = e
 
-            # track the exception on NewRelic but don't stop parsing future
-            # log lines.
-            newrelic.agent.notice_error()
         else:
             completed_names.add(job_log.name)
 

treeherder/middleware.py

@@ -1,6 +1,5 @@
 import re
 
-import newrelic.agent
 from django.utils.deprecation import MiddlewareMixin
 from whitenoise.middleware import WhiteNoiseMiddleware
 
@@ -69,12 +68,3 @@ def immutable_file_test(self, path, url):
         #   bootstrap.min.abda843684d0.js
         return super().immutable_file_test(path, url)
 
-
-class NewRelicMiddleware(MiddlewareMixin):
-    """Adds custom annotations to New Relic web transactions."""
-
-    def process_request(self, request):
-        # The New Relic Python agent only submits the User Agent to APM (for exceptions and
-        # slow transactions), so for use in Insights we have to add it as a customer parameter.
-        if "HTTP_USER_AGENT" in request.META:
-            newrelic.agent.add_custom_attribute("user_agent", request.META["HTTP_USER_AGENT"])

treeherder/model/error_summary.py

@@ -2,7 +2,6 @@
 import logging
 import re
 
-import newrelic.agent
 from django.core.cache import caches
 
 from treeherder.model.models import Bugscache, TextLogError
@@ -174,15 +173,13 @@ def get_error_summary(job, queryset=None):
     try:
         cache.set(cache_key, error_summary, BUG_SUGGESTION_CACHE_TIMEOUT)
     except Exception as e:
-        newrelic.agent.record_custom_event("error caching error_summary for job", job.id)
         logger.error("error caching error_summary for job %s: %s", job.id, e, exc_info=True)
 
     try:
         lcache.update_cache(date, line_cache[date])
         # TODO: consider reducing this, each date is ~5%, so it will be faster
         lcache.update_db_cache(date, line_cache[date])
     except Exception as e:
-        newrelic.agent.record_custom_event("error caching error_lines for job", job.id)
         logger.error("error caching error_lines for job %s: %s", job.id, e, exc_info=True)
 
     return error_summary

treeherder/model/models.py

@@ -5,7 +5,6 @@
 import warnings
 from hashlib import sha1
 
-import newrelic.agent
 from django.conf import settings
 from django.contrib.auth.models import User
 from django.contrib.postgres.indexes import GinIndex
@@ -21,7 +20,6 @@
 
 from treeherder.webapp.api.utils import REPO_GROUPS, to_timestamp
 
-warnings.filterwarnings("ignore", category=DeprecationWarning, module="newrelic")
 
 logger = logging.getLogger(__name__)
 
@@ -294,7 +292,6 @@ def search(cls, search_term):
             open_recent = [x for x in all_data if x["resolution"] == ""]
             all_others = [x for x in all_data if x["resolution"] != ""]
         except ProgrammingError as e:
-            newrelic.agent.notice_error()
             logger.error(
                 f"Failed to execute FULLTEXT search on Bugscache, error={e}, SQL={recent_qs.query.__str__()}"
             )
@@ -1291,19 +1288,11 @@ def verify_classification(self, classification):
             self.metadata.best_is_verified = True
             self.metadata.save(update_fields=["best_classification", "best_is_verified"])
 
-        # Send event to NewRelic when a verifing an autoclassified failure.
+
         match = self.matches.filter(classified_failure=classification).first()
         if not match:
             return
 
-        newrelic.agent.record_custom_event(
-            "user_verified_classification",
-            {
-                "matcher": match.matcher_name,
-                "job_id": self.id,
-            },
-        )
-
     def get_failure_line(self):
         """Get a related FailureLine instance if one exists."""
         try: