Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

fix(release-alert): Update Query to use Org & Project Filtering for Release Health #71386

Closed
Closed
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
126 changes: 125 additions & 1 deletion src/sentry/release_health/release_monitor/metrics.py
Original file line number Diff line number Diff line change
Expand Up @@ -181,7 +181,122 @@ def fetch_projects_with_recent_sessions(self) -> Mapping[int, Sequence[int]]:
return self.fetch_projects_with_recent_sessions_with_filter()
return self.fetch_projects_with_recent_sessions_with_offset()

def fetch_project_release_health_totals(
def fetch_project_release_health_totals_with_org_project_filter(
self, org_id: int, project_ids: Sequence[int]
) -> Totals:
start_time = time.time()
prev_org_id = 0
prev_project_id = 0
totals: Totals = defaultdict(dict)
with metrics.timer("release_monitor.fetch_project_release_health_totals.loop"):
while (time.time() - start_time) < self.MAX_SECONDS:
release_key = resolve_tag_key(UseCaseID.SESSIONS, org_id, "release")
release_col = Column(release_key)
env_key = resolve_tag_key(UseCaseID.SESSIONS, org_id, "environment")
env_col = Column(env_key)
query = Query(
match=Entity(EntityKey.MetricsCounters.value),
select=[
Function("sum", [Column("value")], "sessions"),
Column("project_id"),
release_col,
env_col,
],
groupby=[
Column("org_id"),
Column("project_id"),
release_col,
env_col,
],
where=[
Condition(
Column("timestamp"),
Op.GTE,
datetime.utcnow() - timedelta(hours=6),
),
Condition(Column("timestamp"), Op.LT, datetime.utcnow()),
Condition(Column("org_id"), Op.EQ, org_id),
Condition(Column("project_id"), Op.IN, project_ids),
Condition(
Column("metric_id"),
Op.EQ,
indexer.resolve(
UseCaseID.SESSIONS, org_id, SessionMRI.RAW_SESSION.value
),
),
# Either org_id > prev_org_id or (org_id == prev_org_id and project_id > prev_project_id)
BooleanCondition(
BooleanOp.OR,
[
Condition(Column("org_id"), Op.GT, prev_org_id),
BooleanCondition(
BooleanOp.AND,
[
Condition(Column("org_id"), Op.EQ, prev_org_id),
Condition(Column("project_id"), Op.GT, prev_project_id),
],
),
],
),
],
granularity=Granularity(21600),
orderby=[
OrderBy(Column("org_id"), Direction.ASC),
OrderBy(Column("project_id"), Direction.ASC),
OrderBy(release_col, Direction.ASC),
OrderBy(env_col, Direction.ASC),
],
).set_limit(self.CHUNK_SIZE + 1)
request = Request(
dataset=Dataset.Metrics.value,
app_id="release_health",
query=query,
tenant_ids={"organization_id": org_id},
)
with metrics.timer("release_monitor.fetch_project_release_health_totals.query"):
data = raw_snql_query(
request, "release_monitor.fetch_project_release_health_totals"
)["data"]
count = len(data)
more_results = count > self.CHUNK_SIZE

if more_results:
data = data[:-1]

# convert indexes back to strings
indexes: set[int] = set()
for row in data:
indexes.add(row[env_key])
indexes.add(row[release_key])
resolved_strings = indexer.bulk_reverse_resolve(
UseCaseID.SESSIONS, org_id, indexes
)

for row in data:
env_name = resolved_strings.get(row[env_key])
release_name = resolved_strings.get(row[release_key])
row_totals = totals[row["project_id"]].setdefault(
env_name, {"total_sessions": 0, "releases": defaultdict(int)} # type: ignore[arg-type]
)
row_totals["total_sessions"] += row["sessions"]
row_totals["releases"][release_name] += row["sessions"] # type: ignore[index]

# Update prev_org_id and prev_project_id for the next iteration
if data:
last_row = data[-1]
prev_org_id = last_row["org_id"]
prev_project_id = last_row["project_id"]

if not more_results:
break
else:
logger.error(
"fetch_project_release_health_totals.loop_timeout",
extra={"org_id": org_id, "project_ids": project_ids},
)
return totals

def fetch_project_release_health_totals_with_offset(
self, org_id: int, project_ids: Sequence[int]
) -> Totals:
start_time = time.time()
Expand Down Expand Up @@ -277,3 +392,12 @@ def fetch_project_release_health_totals(
extra={"org_id": org_id, "project_ids": project_ids},
)
return totals

def fetch_project_release_health_totals(
self, org_id: int, project_ids: Sequence[int]
) -> Totals:
if options.get("release-health.use-org-and-project-filter"):
return self.fetch_project_release_health_totals_with_org_project_filter(
org_id, project_ids
)
return self.fetch_project_release_health_totals_with_offset(org_id, project_ids)
41 changes: 40 additions & 1 deletion tests/sentry/release_health/release_monitor/__init__.py
Original file line number Diff line number Diff line change
Expand Up @@ -89,7 +89,46 @@ def setUp(self):
self.release2 = self.create_release(project=self.project2)
self.backend = self.backend_class()

def test(self):
def test_with_offset(self):
self.bulk_store_sessions(
[
self.build_session(
project_id=self.project1,
environment=self.environment1.name,
release=self.release1.version,
)
for _ in range(5)
]
+ [
self.build_session(
project_id=self.project2,
environment=self.environment2.name,
release=self.release2.version,
)
]
)

totals = self.backend.fetch_project_release_health_totals(
self.organization.id,
[self.project.id, self.project1.id, self.project2.id],
)
assert totals == {
self.project1.id: {
self.environment1.name: {
"total_sessions": 5,
"releases": {self.release1.version: 5},
}
},
self.project2.id: {
self.environment2.name: {
"total_sessions": 1,
"releases": {self.release2.version: 1},
}
},
}, totals

@override_options({"release-health.use-org-and-project-filter": True})
def test_with_filter(self):
self.bulk_store_sessions(
[
self.build_session(
Expand Down
Loading