Xray
daily rollups 1h 6h 24h 48h 3d 5d 7d 14d 30d ‹
Last 30 days▾
Ends now. Anything up to this cluster's 30-day retention.
Start and end times
Cancel
Interpreted in UTC (change it at the right of this bar). Earliest available: 2026-10-19 12:07:39.
› →|
updated 1d ago
Sample data. This database was generated by scripts/seed_dev_db.py — every service, deploy, person and figure on this page is synthetic. Names are sample.*@seed.invalid and image tags are prefixed SEED-. Nothing here happened.
← findings

feedsvc shape 00000000: p95 4.85 s now vs 28 ms earlier (172.7×, 3,394 runs)

warning Query shape degrading test26 resolved impact 30 opened 20d ago · last observed 1d ago · observed 14× · #8769 · fd21d23b22050dcbf624

What happened template

A feedsvc query shape on test26 has p95 4.8 s in the second half of the window against 28 ms in the first (172.7×, 3,394 executions). The shape reads: “RETURN EXISTS { MATCH (cfg2:WorkplanCfg) WHERE cfg2.progressPending = true AND coalesce(cfg2.lastViewedAtMs, ?) > (times…”. Slower can mean more data, a lost index, or lock contention — the shape page has the CPU/wait split. Shapes whose p95 stays under a second are never reported: below that nobody notices, whatever the ratio.

A Cypher shape whose p95 is worsening across the window and is at least 1 s now; spikes under a second are ignored.

The facts

servicefeedsvc
plan_hash00000000
queries3,394
floor_ms1,000
p95_early28.09
p95_late4,850
ratio172.69
avg_ms9.58
sampleRETURN EXISTS { MATCH (cfg2:WorkplanCfg) WHERE cfg2.progressPending = true AND coalesce(cfg2.lastViewedAtMs, ?) > (timestamp() - ?) AND (timestamp() - coalesce(cfg2.lastProgressComputedAtMs, ?)) > coalesce(cfg2.progressComputeIntervalMs, ?) } AS fleetProgressPending, EXISTS { MATCH (item:WorkplanInstanceMilestoneItem) WHERE item.itemType = ? AND coalesce(item.workOrderId, ?) <> ? AND coalesce(item
rwwrite
severity_cappedcritical → warning: test26 is a test-tier cluster
discriminator00000000
window2026-11-17 03:59:37 UTC → 2026-11-17 09:59:37 UTC

The story — No feedsvc deploy or hand change is recorded in the hour before this opened. Everything below is feedsvc only.

History

  • 2026-11-17 09:22:05 UTCreopened x-ray · severity=warning was=warning title=feedsvc shape 00000000: p95 4.85 s now vs 28 ms earlier (173.5×, 4,656 runs) level=low was_level=low finding={'id': 8769, 'fingerprint': 'fd21d23b22050dcbf624', 'kind': 'shape.degrading', 'severity': 'warning', 'cluster': 'test26', 'subject': 'feedsvc', 'title': 'feedsvc shape 00000000: p95 4.85 s now vs 28 ms earlier (173.5×, 4,656 runs)', 'facts': {'service': 'feedsvc', 'plan_hash': '00000000', 'queries': 4656, 'floor_ms': 1000, 'p95_early': 27.95731707317073, 'p95_late': 4850.0, 'ratio': 173.47873500545256, 'avg_ms': 8.210051546391753, 'sample': 'MATCH (currentUser:User{deltimestamp:?,deltimestamp:?,email:$email}) WHERE (currentUser.isDummy IS NULL OR currentUser.isDummy <> true) OPTIONAL MATCH (supportUser:User{email:$supportUserEmail}) OPTIONAL MATCH (targetOrg:Organization{delpending:false, alias:$orgAlias}) // A deactivated (Inactive) org actor is not a membership (ENG-?). OPTIONAL MATCH (currentUser)-[:FOR_USER]-(orgActor:Actor{delpen', 'rw': 'write', 'window': [1794885725, 1794907325], 'recent_deploys': [], 'severity_capped': 'critical → warning: test26 is a test-tier cluster', 'discriminator': '00000000'}, 'evidence': [{'label': 'Query shape', 'href': '/neo/shape/00000000?cluster=test26&from=1794885725&to=1794907325'}, {'label': 'Degrading shapes', 'href': '/neo/queries?cluster=test26&from=1794885725&to=1794907325&order=degrading'}], 'provenance': {'source': 'rollup', 'collected_at': 1794907321, 'age_seconds': 13, 'stale': False, 'shown': None, 'total': None, 'truncated': False, 'coverage_from': None, 'retain_days': None, 'note': None}, 'impact': 24.7, 'status': 'open', 'first_seen': 1793210755, 'last_seen': 1794907325, 'resolved_at': None, 'acked_by': None, 'acked_at': None, 'snoozed_until': None, 'occurrences': 12, 'narrative': 'A feedsvc query shape on test26 has p95 1.8 s in the second half of the window against 9 ms in the first (198.3×, 280 executions). The shape reads: “MATCH (task:Task) WHERE task.task_number = ? AND EXISTS { (task)-[:BELONGS_TO]->(:Inbox{type:?})-[:MANAGED_BY]->(:Organi…”. Slower can mean more data, a lost index, or lock contention — the shape page has the CPU/wait split. Shapes whose p95 stays under a second are never reported: below that nobody notices, whatever the ratio.', 'narrative_kind': 'template', 'discriminator': '00000000', 'level': 'low', 'told_level': None, 'episode_at': 1794907325, 'chat_thread': None, 'kind_label': 'Query shape degrading', 'group': 'Neo4j', 'age_s': 1696578, 'stale_s': 8}
  • 2026-11-17 10:21:45 UTCresolved x-ray · reason=not observed grace_s=900 finding={'id': 8769, 'fingerprint': 'fd21d23b22050dcbf624', 'kind': 'shape.degrading', 'severity': 'warning', 'cluster': 'test26', 'subject': 'feedsvc', 'title': 'feedsvc shape 00000000: p95 4.85 s now vs 28 ms earlier (172.7×, 3,394 runs)', 'facts': {'service': 'feedsvc', 'plan_hash': '00000000', 'queries': 3394, 'floor_ms': 1000, 'p95_early': 28.085561497326147, 'p95_late': 4850.0, 'ratio': 172.68659558263553, 'avg_ms': 9.579846788450206, 'sample': 'RETURN EXISTS { MATCH (cfg2:WorkplanCfg) WHERE cfg2.progressPending = true AND coalesce(cfg2.lastViewedAtMs, ?) > (timestamp() - ?) AND (timestamp() - coalesce(cfg2.lastProgressComputedAtMs, ?)) > coalesce(cfg2.progressComputeIntervalMs, ?) } AS fleetProgressPending, EXISTS { MATCH (item:WorkplanInstanceMilestoneItem) WHERE item.itemType = ? AND coalesce(item.workOrderId, ?) <> ? AND coalesce(item', 'rw': 'write', 'window': [1794887977, 1794909577], 'recent_deploys': [], 'severity_capped': 'critical → warning: test26 is a test-tier cluster', 'discriminator': '00000000'}, 'evidence': [{'label': 'Query shape', 'href': '/neo/shape/00000000?cluster=test26&from=1794887977&to=1794909577'}, {'label': 'Degrading shapes', 'href': '/neo/queries?cluster=test26&from=1794887977&to=1794909577&order=degrading'}], 'provenance': {'source': 'rollup', 'collected_at': 1794909577, 'age_seconds': 9, 'stale': False, 'shown': None, 'total': None, 'truncated': False, 'coverage_from': None, 'retain_days': None, 'note': None}, 'impact': 29.7, 'status': 'resolved', 'first_seen': 1793210755, 'last_seen': 1794909577, 'resolved_at': 1794910905, 'acked_by': None, 'acked_at': None, 'snoozed_until': None, 'occurrences': 14, 'narrative': 'A feedsvc query shape on test26 has p95 4.8 s in the second half of the window against 28 ms in the first (172.7×, 3,394 executions). The shape reads: “RETURN EXISTS { MATCH (cfg2:WorkplanCfg) WHERE cfg2.progressPending = true AND coalesce(cfg2.lastViewedAtMs, ?) > (times…”. Slower can mean more data, a lost index, or lock contention — the shape page has the CPU/wait split. Shapes whose p95 stays under a second are never reported: below that nobody notices, whatever the ratio.', 'narrative_kind': 'template', 'discriminator': '00000000', 'level': 'low', 'told_level': None, 'episode_at': 1794907325, 'chat_thread': None, 'kind_label': 'Query shape degrading', 'group': 'Neo4j', 'age_s': 1700158, 'stale_s': 1336}

Check it yourself

Every link carries the finding's own time window, so it shows tomorrow what it shows now.

Actions

Sign in (IAP) to acknowledge, snooze or resolve.

Where the numbers come from

sourcerollup
collected_at1794909577
age_seconds9
staleFalse
shownNone
totalNone
truncatedFalse
coverage_fromNone
retain_daysNone
noteNone
detectorshapes