studiosvc p95 26.46 s vs a normal 485 ms ± 437 ms for Tue 10:00 UTC (3 of the last 4 checks)
info
Latency anomaly
automation
open
impact 13
Owner: Suresh Boddapati (backup: Madhava Mukkamala)
opened 15d ago · last observed 1d ago · observed 198× · #15924 · 2cc6f9770ec51367864a
What happened template
studiosvc's p95 is 26.5 s on automation, above a normal of 485 ms ± 437 ms for this hour of the week (3 of the last 4 five-minute buckets, z = 59.4, 3 weeks of history). Whether users noticed depends on which endpoints carried it — the service page shows the split. A p95 that stays under a second is never reported — below that nobody notices, whatever the deviation.
p95 far above its hour-of-week normal — and at least 1 s; under a second is noise.
The facts
| measure | api_p95_ms |
| unit | ms |
| value | 26,462 |
| expected_median | 485.00 |
| expected_mad | 437.47 |
| z | 59.38 |
| rel | 53.56 |
| buckets_hit | 3 |
| n | 28 |
| weeks | 3 |
| slot | Tue 10:00 UTC |
| bucket | 2026-11-17 10:10:00 UTC (1d ago) |
| floor_ms | 1,000 |
| discriminator | api_p95_ms |
| window | 2026-11-17 10:00:00 UTC → 2026-11-17 10:20:00 UTC |
What is failing in studiosvc — the bad events, decomposed
“Bad events” are HTTP 5xx answers from studiosvc. In this finding's window (2026-11-17 10:00:00 UTC → 2026-11-17 10:20:00 UTC) there were 0; every line links to the grouped errors behind it.
The error shapes behind them
| 1403× | http error response returned to client since 2026-10-07 10:27:55 UTC |
| 5× | Request completed — 500 · could not create activation rule chain:activation rule \ since 2026-09-25 09:22:48 UTC |
History
- 2026-11-14 20:13:09 UTCresolved x-ray · reason=not observed grace_s=900 finding={'id': 15924, 'fingerprint': '2cc6f9770ec51367864a', 'kind': 'anomaly.latency', 'severity': 'info', 'cluster': 'automation', 'subject': 'studiosvc', 'title': 'studiosvc p95 1.9 min vs a normal 485 ms ± 474 ms for Sat 09:00 UTC (4 of the last 4 checks)', 'facts': {'measure': 'api_p95_ms', 'unit': 'ms', 'value': 117000.0, 'expected_median': 485.0, 'expected_mad': 474.432, 'z': 245.5884088762984, 'rel': 240.23711340206185, 'buckets_hit': 4, 'n': 21, 'weeks': 2, 'slot': 'Sat 09:00 UTC', 'bucket': 1794650100, 'window': [1794649200, 1794650400], 'floor_ms': 1000, 'recent_deploys': [], 'discriminator': 'api_p95_ms'}, 'evidence': [{'label': 'Service', 'href': '/service/studiosvc?cluster=automation&from=1794627600&to=1794650400'}, {'label': 'Fleet vs normal', 'href': '/fleet?cluster=automation&from=1794627600&to=1794650400'}], 'provenance': {'source': 'rollup', 'collected_at': 1794650895, 'age_seconds': 44, 'stale': False, 'shown': None, 'total': None, 'truncated': False, 'coverage_from': None, 'retain_days': None, 'note': None}, 'impact': 12.6, 'status': 'resolved', 'first_seen': 1793677716, 'last_seen': 1794650932, 'resolved_at': 1794687189, 'acked_by': None, 'acked_at': None, 'snoozed_until': None, 'occurrences': 145, 'narrative': "studiosvc's p95 is 117.0 s on automation, above a normal of 485 ms ± 474 ms for this hour of the week (4 of the last 4 five-minute buckets, z = 245.6, 2 weeks of history). Whether users noticed depends on which endpoints carried it — the service page shows the split. A p95 that stays under a second is never reported — below that nobody notices, whatever the deviation.", 'narrative_kind': 'template', 'discriminator': 'api_p95_ms', 'level': 'low', 'told_level': None, 'episode_at': 1794646792, 'chat_thread': None, 'kind_label': 'Latency anomaly', 'group': 'Anomalies', 'age_s': 1009484, 'stale_s': 36268}
- 2026-11-15 02:13:12 UTCreopened x-ray · severity=info was=info title=studiosvc p95 2.0 min vs a normal 195 ms ± 170 ms for Sun 01:00 UTC (3 of the last 4 checks) level=low was_level=low finding={'id': 15924, 'fingerprint': '2cc6f9770ec51367864a', 'kind': 'anomaly.latency', 'severity': 'info', 'cluster': 'automation', 'subject': 'studiosvc', 'title': 'studiosvc p95 2.0 min vs a normal 195 ms ± 170 ms for Sun 01:00 UTC (3 of the last 4 checks)', 'facts': {'measure': 'api_p95_ms', 'unit': 'ms', 'value': 120000.0, 'expected_median': 195.0, 'expected_mad': 170.499, 'z': 702.672742948639, 'rel': 614.3846153846154, 'buckets_hit': 3, 'n': 41, 'weeks': 4, 'slot': 'Sun 01:00 UTC', 'bucket': 1794707700, 'window': [1794706800, 1794708000], 'floor_ms': 1000, 'recent_deploys': [], 'discriminator': 'api_p95_ms'}, 'evidence': [{'label': 'Service', 'href': '/service/studiosvc?cluster=automation&from=1794685200&to=1794708000'}, {'label': 'Fleet vs normal', 'href': '/fleet?cluster=automation&from=1794685200&to=1794708000'}], 'provenance': {'source': 'rollup', 'collected_at': 1794708579, 'age_seconds': 221, 'stale': False, 'shown': None, 'total': None, 'truncated': False, 'coverage_from': None, 'retain_days': None, 'note': None}, 'impact': 7.2, 'status': 'open', 'first_seen': 1793677716, 'last_seen': 1794708792, 'resolved_at': None, 'acked_by': None, 'acked_at': None, 'snoozed_until': None, 'occurrences': 146, 'narrative': "studiosvc's p95 is 117.0 s on automation, above a normal of 485 ms ± 474 ms for this hour of the week (4 of the last 4 five-minute buckets, z = 245.6, 2 weeks of history). Whether users noticed depends on which endpoints carried it — the service page shows the split. A p95 that stays under a second is never reported — below that nobody notices, whatever the deviation.", 'narrative_kind': 'template', 'discriminator': 'api_p95_ms', 'level': 'low', 'told_level': None, 'episode_at': 1794708792, 'chat_thread': None, 'kind_label': 'Latency anomaly', 'group': 'Anomalies', 'age_s': 1031084, 'stale_s': 8}
- 2026-11-15 19:38:48 UTCresolved x-ray · reason=not observed grace_s=900 finding={'id': 15924, 'fingerprint': '2cc6f9770ec51367864a', 'kind': 'anomaly.latency', 'severity': 'info', 'cluster': 'automation', 'subject': 'studiosvc', 'title': 'studiosvc p95 2.0 min vs a normal 192 ms ± 271 ms for Sun 15:00 UTC (4 of the last 4 checks)', 'facts': {'measure': 'api_p95_ms', 'unit': 'ms', 'value': 120000.0, 'expected_median': 192.5, 'expected_mad': 270.94516, 'z': 442.1835769275229, 'rel': 622.3766233766233, 'buckets_hit': 4, 'n': 25, 'weeks': 3, 'slot': 'Sun 15:00 UTC', 'bucket': 1794757500, 'window': [1794756600, 1794757800], 'floor_ms': 1000, 'recent_deploys': [], 'discriminator': 'api_p95_ms'}, 'evidence': [{'label': 'Service', 'href': '/service/studiosvc?cluster=automation&from=1794735000&to=1794757800'}, {'label': 'Fleet vs normal', 'href': '/fleet?cluster=automation&from=1794735000&to=1794757800'}], 'provenance': {'source': 'rollup', 'collected_at': 1794758316, 'age_seconds': 38, 'stale': False, 'shown': None, 'total': None, 'truncated': False, 'coverage_from': None, 'retain_days': None, 'note': None}, 'impact': 12.2, 'status': 'resolved', 'first_seen': 1793677716, 'last_seen': 1794758345, 'resolved_at': 1794771528, 'acked_by': None, 'acked_at': None, 'snoozed_until': None, 'occurrences': 171, 'narrative': "studiosvc's p95 is 120.0 s on automation, above a normal of 192 ms ± 271 ms for this hour of the week (4 of the last 4 five-minute buckets, z = 442.2, 3 weeks of history). Whether users noticed depends on which endpoints carried it — the service page shows the split. A p95 that stays under a second is never reported — below that nobody notices, whatever the deviation.", 'narrative_kind': 'template', 'discriminator': 'api_p95_ms', 'level': 'low', 'told_level': None, 'episode_at': 1794708792, 'chat_thread': None, 'kind_label': 'Latency anomaly', 'group': 'Anomalies', 'age_s': 1093822, 'stale_s': 13193}
- 2026-11-16 15:08:36 UTCreopened x-ray · severity=info was=info title=studiosvc p95 2.0 min vs a normal 195 ms ± 77 ms for Mon 14:00 UTC (3 of the last 4 checks) level=low was_level=low finding={'id': 15924, 'fingerprint': '2cc6f9770ec51367864a', 'kind': 'anomaly.latency', 'severity': 'info', 'cluster': 'automation', 'subject': 'studiosvc', 'title': 'studiosvc p95 2.0 min vs a normal 195 ms ± 77 ms for Mon 14:00 UTC (3 of the last 4 checks)', 'facts': {'measure': 'api_p95_ms', 'unit': 'ms', 'value': 120000.0, 'expected_median': 195.0, 'expected_mad': 76.601, 'z': 1564.0135246276159, 'rel': 614.3846153846154, 'buckets_hit': 3, 'n': 22, 'weeks': 5, 'slot': 'Mon 14:00 UTC', 'bucket': 1794840900, 'window': [1794840000, 1794841200], 'floor_ms': 1000, 'recent_deploys': [], 'discriminator': 'api_p95_ms'}, 'evidence': [{'label': 'Service', 'href': '/service/studiosvc?cluster=automation&from=1794818400&to=1794841200'}, {'label': 'Fleet vs normal', 'href': '/fleet?cluster=automation&from=1794818400&to=1794841200'}], 'provenance': {'source': 'rollup', 'collected_at': 1794841609, 'age_seconds': 120, 'stale': False, 'shown': None, 'total': None, 'truncated': False, 'coverage_from': None, 'retain_days': None, 'note': None}, 'impact': 7.6, 'status': 'open', 'first_seen': 1793677716, 'last_seen': 1794841716, 'resolved_at': None, 'acked_by': None, 'acked_at': None, 'snoozed_until': None, 'occurrences': 172, 'narrative': "studiosvc's p95 is 120.0 s on automation, above a normal of 192 ms ± 271 ms for this hour of the week (4 of the last 4 five-minute buckets, z = 442.2, 3 weeks of history). Whether users noticed depends on which endpoints carried it — the service page shows the split. A p95 that stays under a second is never reported — below that nobody notices, whatever the deviation.", 'narrative_kind': 'template', 'discriminator': 'api_p95_ms', 'level': 'low', 'told_level': None, 'episode_at': 1794841716, 'chat_thread': None, 'kind_label': 'Latency anomaly', 'group': 'Anomalies', 'age_s': 1164013, 'stale_s': 13}
- 2026-11-16 16:07:15 UTCresolved x-ray · reason=not observed grace_s=900 finding={'id': 15924, 'fingerprint': '2cc6f9770ec51367864a', 'kind': 'anomaly.latency', 'severity': 'info', 'cluster': 'automation', 'subject': 'studiosvc', 'title': 'studiosvc p95 2.0 min vs a normal 195 ms ± 77 ms for Mon 14:00 UTC (3 of the last 4 checks)', 'facts': {'measure': 'api_p95_ms', 'unit': 'ms', 'value': 120000.0, 'expected_median': 195.0, 'expected_mad': 76.601, 'z': 1564.0135246276159, 'rel': 614.3846153846154, 'buckets_hit': 3, 'n': 22, 'weeks': 5, 'slot': 'Mon 14:00 UTC', 'bucket': 1794840900, 'window': [1794840000, 1794841200], 'floor_ms': 1000, 'recent_deploys': [], 'discriminator': 'api_p95_ms'}, 'evidence': [{'label': 'Service', 'href': '/service/studiosvc?cluster=automation&from=1794818400&to=1794841200'}, {'label': 'Fleet vs normal', 'href': '/fleet?cluster=automation&from=1794818400&to=1794841200'}], 'provenance': {'source': 'rollup', 'collected_at': 1794841609, 'age_seconds': 120, 'stale': False, 'shown': None, 'total': None, 'truncated': False, 'coverage_from': None, 'retain_days': None, 'note': None}, 'impact': 12.6, 'status': 'resolved', 'first_seen': 1793677716, 'last_seen': 1794841716, 'resolved_at': 1794845235, 'acked_by': None, 'acked_at': None, 'snoozed_until': None, 'occurrences': 172, 'narrative': "studiosvc's p95 is 120.0 s on automation, above a normal of 195 ms ± 77 ms for this hour of the week (3 of the last 4 five-minute buckets, z = 1564.0, 5 weeks of history). Whether users noticed depends on which endpoints carried it — the service page shows the split. A p95 that stays under a second is never reported — below that nobody notices, whatever the deviation.", 'narrative_kind': 'template', 'discriminator': 'api_p95_ms', 'level': 'low', 'told_level': None, 'episode_at': 1794841716, 'chat_thread': None, 'kind_label': 'Latency anomaly', 'group': 'Anomalies', 'age_s': 1167529, 'stale_s': 3529}
- 2026-11-16 16:55:04 UTCreopened x-ray · severity=info was=info title=studiosvc p95 2.0 min vs a normal 190 ms ± 82 ms for Mon 16:00 UTC (3 of the last 4 checks) level=low was_level=low finding={'id': 15924, 'fingerprint': '2cc6f9770ec51367864a', 'kind': 'anomaly.latency', 'severity': 'info', 'cluster': 'automation', 'subject': 'studiosvc', 'title': 'studiosvc p95 2.0 min vs a normal 190 ms ± 82 ms for Mon 16:00 UTC (3 of the last 4 checks)', 'facts': {'measure': 'api_p95_ms', 'unit': 'ms', 'value': 120000.0, 'expected_median': 190.0, 'expected_mad': 81.543, 'z': 1469.2861435071065, 'rel': 630.578947368421, 'buckets_hit': 3, 'n': 25, 'weeks': 3, 'slot': 'Mon 16:00 UTC', 'bucket': 1794847200, 'window': [1794846300, 1794847500], 'floor_ms': 1000, 'recent_deploys': [], 'discriminator': 'api_p95_ms'}, 'evidence': [{'label': 'Service', 'href': '/service/studiosvc?cluster=automation&from=1794824700&to=1794847500'}, {'label': 'Fleet vs normal', 'href': '/fleet?cluster=automation&from=1794824700&to=1794847500'}], 'provenance': {'source': 'rollup', 'collected_at': 1794847901, 'age_seconds': 217, 'stale': False, 'shown': None, 'total': None, 'truncated': False, 'coverage_from': None, 'retain_days': None, 'note': None}, 'impact': 7.7, 'status': 'open', 'first_seen': 1793677716, 'last_seen': 1794848104, 'resolved_at': None, 'acked_by': None, 'acked_at': None, 'snoozed_until': None, 'occurrences': 173, 'narrative': "studiosvc's p95 is 120.0 s on automation, above a normal of 195 ms ± 77 ms for this hour of the week (3 of the last 4 five-minute buckets, z = 1564.0, 5 weeks of history). Whether users noticed depends on which endpoints carried it — the service page shows the split. A p95 that stays under a second is never reported — below that nobody notices, whatever the deviation.", 'narrative_kind': 'template', 'discriminator': 'api_p95_ms', 'level': 'low', 'told_level': None, 'episode_at': 1794848104, 'chat_thread': None, 'kind_label': 'Latency anomaly', 'group': 'Anomalies', 'age_s': 1170402, 'stale_s': 14}
- 2026-11-16 17:27:13 UTCresolved x-ray · reason=not observed grace_s=900 finding={'id': 15924, 'fingerprint': '2cc6f9770ec51367864a', 'kind': 'anomaly.latency', 'severity': 'info', 'cluster': 'automation', 'subject': 'studiosvc', 'title': 'studiosvc p95 19.40 s vs a normal 190 ms ± 82 ms for Mon 16:00 UTC (3 of the last 4 checks)', 'facts': {'measure': 'api_p95_ms', 'unit': 'ms', 'value': 19399.999999999993, 'expected_median': 190.0, 'expected_mad': 81.543, 'z': 235.5812270826434, 'rel': 101.1052631578947, 'buckets_hit': 3, 'n': 25, 'weeks': 3, 'slot': 'Mon 16:00 UTC', 'bucket': 1794848100, 'window': [1794847200, 1794848400], 'floor_ms': 1000, 'recent_deploys': [], 'discriminator': 'api_p95_ms'}, 'evidence': [{'label': 'Service', 'href': '/service/studiosvc?cluster=automation&from=1794825600&to=1794848400'}, {'label': 'Fleet vs normal', 'href': '/fleet?cluster=automation&from=1794825600&to=1794848400'}], 'provenance': {'source': 'rollup', 'collected_at': 1794848776, 'age_seconds': 164, 'stale': False, 'shown': None, 'total': None, 'truncated': False, 'coverage_from': None, 'retain_days': None, 'note': None}, 'impact': 12.9, 'status': 'resolved', 'first_seen': 1793677716, 'last_seen': 1794848928, 'resolved_at': 1794850033, 'acked_by': None, 'acked_at': None, 'snoozed_until': None, 'occurrences': 174, 'narrative': "studiosvc's p95 is 19.4 s on automation, above a normal of 190 ms ± 82 ms for this hour of the week (3 of the last 4 five-minute buckets, z = 235.6, 3 weeks of history). Whether users noticed depends on which endpoints carried it — the service page shows the split. A p95 that stays under a second is never reported — below that nobody notices, whatever the deviation.", 'narrative_kind': 'template', 'discriminator': 'api_p95_ms', 'level': 'low', 'told_level': None, 'episode_at': 1794848104, 'chat_thread': None, 'kind_label': 'Latency anomaly', 'group': 'Anomalies', 'age_s': 1172328, 'stale_s': 1116}
- 2026-11-17 02:14:26 UTCreopened x-ray · severity=info was=info title=studiosvc p95 1.7 min vs a normal 409 ms ± 318 ms for Tue 02:00 UTC (3 of the last 4 checks) level=low was_level=low finding={'id': 15924, 'fingerprint': '2cc6f9770ec51367864a', 'kind': 'anomaly.latency', 'severity': 'info', 'cluster': 'automation', 'subject': 'studiosvc', 'title': 'studiosvc p95 1.7 min vs a normal 409 ms ± 318 ms for Tue 02:00 UTC (3 of the last 4 checks)', 'facts': {'measure': 'api_p95_ms', 'unit': 'ms', 'value': 102230.76923076919, 'expected_median': 409.34784, 'expected_mad': 317.79208, 'z': 320.40263996122616, 'rel': 248.74058549025, 'buckets_hit': 3, 'n': 51, 'weeks': 5, 'slot': 'Tue 02:00 UTC', 'bucket': 1794880800, 'window': [1794879900, 1794881100], 'floor_ms': 1000, 'recent_deploys': [], 'discriminator': 'api_p95_ms'}, 'evidence': [{'label': 'Service', 'href': '/service/studiosvc?cluster=automation&from=1794858300&to=1794881100'}, {'label': 'Fleet vs normal', 'href': '/fleet?cluster=automation&from=1794858300&to=1794881100'}], 'provenance': {'source': 'rollup', 'collected_at': 1794881576, 'age_seconds': 102, 'stale': False, 'shown': None, 'total': None, 'truncated': False, 'coverage_from': None, 'retain_days': None, 'note': None}, 'impact': 7.3, 'status': 'open', 'first_seen': 1793677716, 'last_seen': 1794881666, 'resolved_at': None, 'acked_by': None, 'acked_at': None, 'snoozed_until': None, 'occurrences': 175, 'narrative': "studiosvc's p95 is 19.4 s on automation, above a normal of 190 ms ± 82 ms for this hour of the week (3 of the last 4 five-minute buckets, z = 235.6, 3 weeks of history). Whether users noticed depends on which endpoints carried it — the service page shows the split. A p95 that stays under a second is never reported — below that nobody notices, whatever the deviation.", 'narrative_kind': 'template', 'discriminator': 'api_p95_ms', 'level': 'low', 'told_level': None, 'episode_at': 1794881666, 'chat_thread': None, 'kind_label': 'Latency anomaly', 'group': 'Anomalies', 'age_s': 1203963, 'stale_s': 13}
- 2026-11-17 04:45:57 UTCresolved x-ray · reason=not observed grace_s=900 finding={'id': 15924, 'fingerprint': '2cc6f9770ec51367864a', 'kind': 'anomaly.latency', 'severity': 'info', 'cluster': 'automation', 'subject': 'studiosvc', 'title': 'studiosvc p95 1.8 min vs a normal 470 ms ± 556 ms for Tue 03:00 UTC (3 of the last 4 checks)', 'facts': {'measure': 'api_p95_ms', 'unit': 'ms', 'value': 110666.66666666663, 'expected_median': 470.0, 'expected_mad': 555.975, 'z': 198.20435571143778, 'rel': 234.46099290780134, 'buckets_hit': 3, 'n': 39, 'weeks': 5, 'slot': 'Tue 03:00 UTC', 'bucket': 1794887700, 'window': [1794887100, 1794888300], 'floor_ms': 1000, 'recent_deploys': [], 'discriminator': 'api_p95_ms'}, 'evidence': [{'label': 'Service', 'href': '/service/studiosvc?cluster=automation&from=1794865500&to=1794888300'}, {'label': 'Fleet vs normal', 'href': '/fleet?cluster=automation&from=1794865500&to=1794888300'}], 'provenance': {'source': 'rollup', 'collected_at': 1794888885, 'age_seconds': 118, 'stale': False, 'shown': None, 'total': None, 'truncated': False, 'coverage_from': None, 'retain_days': None, 'note': None}, 'impact': 12.4, 'status': 'resolved', 'first_seen': 1793677716, 'last_seen': 1794888990, 'resolved_at': 1794890757, 'acked_by': None, 'acked_at': None, 'snoozed_until': None, 'occurrences': 186, 'narrative': "studiosvc's p95 is 110.7 s on automation, above a normal of 470 ms ± 556 ms for this hour of the week (3 of the last 4 five-minute buckets, z = 198.2, 5 weeks of history). Whether users noticed depends on which endpoints carried it — the service page shows the split. A p95 that stays under a second is never reported — below that nobody notices, whatever the deviation.", 'narrative_kind': 'template', 'discriminator': 'api_p95_ms', 'level': 'low', 'told_level': None, 'episode_at': 1794881666, 'chat_thread': None, 'kind_label': 'Latency anomaly', 'group': 'Anomalies', 'age_s': 1213055, 'stale_s': 1781}
- 2026-11-17 06:29:15 UTCreopened x-ray · severity=info was=info title=studiosvc p95 23.37 s vs a normal 1.30 s ± 1.78 s for Tue 06:00 UTC (3 of the last 4 checks) level=low was_level=low finding={'id': 15924, 'fingerprint': '2cc6f9770ec51367864a', 'kind': 'anomaly.latency', 'severity': 'info', 'cluster': 'automation', 'subject': 'studiosvc', 'title': 'studiosvc p95 23.37 s vs a normal 1.30 s ± 1.78 s for Tue 06:00 UTC (3 of the last 4 checks)', 'facts': {'measure': 'api_p95_ms', 'unit': 'ms', 'value': 23374.999999999985, 'expected_median': 1300.0, 'expected_mad': 1782.8265, 'z': 12.382023713468465, 'rel': 16.98076923076922, 'buckets_hit': 3, 'n': 29, 'weeks': 4, 'slot': 'Tue 06:00 UTC', 'bucket': 1794895800, 'window': [1794894900, 1794896100], 'floor_ms': 1000, 'recent_deploys': [], 'discriminator': 'api_p95_ms'}, 'evidence': [{'label': 'Service', 'href': '/service/studiosvc?cluster=automation&from=1794873300&to=1794896100'}, {'label': 'Fleet vs normal', 'href': '/fleet?cluster=automation&from=1794873300&to=1794896100'}], 'provenance': {'source': 'rollup', 'collected_at': 1794896645, 'age_seconds': 322, 'stale': False, 'shown': None, 'total': None, 'truncated': False, 'coverage_from': None, 'retain_days': None, 'note': None}, 'impact': 7.6, 'status': 'open', 'first_seen': 1793677716, 'last_seen': 1794896955, 'resolved_at': None, 'acked_by': None, 'acked_at': None, 'snoozed_until': None, 'occurrences': 187, 'narrative': "studiosvc's p95 is 110.7 s on automation, above a normal of 470 ms ± 556 ms for this hour of the week (3 of the last 4 five-minute buckets, z = 198.2, 5 weeks of history). Whether users noticed depends on which endpoints carried it — the service page shows the split. A p95 that stays under a second is never reported — below that nobody notices, whatever the deviation.", 'narrative_kind': 'template', 'discriminator': 'api_p95_ms', 'level': 'low', 'told_level': None, 'episode_at': 1794896955, 'chat_thread': None, 'kind_label': 'Latency anomaly', 'group': 'Anomalies', 'age_s': 1219251, 'stale_s': 12}
- 2026-11-17 08:55:54 UTCresolved x-ray · reason=not observed grace_s=900 finding={'id': 15924, 'fingerprint': '2cc6f9770ec51367864a', 'kind': 'anomaly.latency', 'severity': 'info', 'cluster': 'automation', 'subject': 'studiosvc', 'title': 'studiosvc p95 23.87 s vs a normal 489 ms ± 446 ms for Tue 08:00 UTC (4 of the last 4 checks)', 'facts': {'measure': 'api_p95_ms', 'unit': 'ms', 'value': 23874.999999999956, 'expected_median': 488.75, 'expected_mad': 446.0155, 'z': 52.43371586861882, 'rel': 47.84910485933495, 'buckets_hit': 4, 'n': 34, 'weeks': 5, 'slot': 'Tue 08:00 UTC', 'bucket': 1794902400, 'window': [1794901500, 1794902700], 'floor_ms': 1000, 'recent_deploys': [], 'discriminator': 'api_p95_ms'}, 'evidence': [{'label': 'Service', 'href': '/service/studiosvc?cluster=automation&from=1794879900&to=1794902700'}, {'label': 'Fleet vs normal', 'href': '/fleet?cluster=automation&from=1794879900&to=1794902700'}], 'provenance': {'source': 'rollup', 'collected_at': 1794903237, 'age_seconds': 779, 'stale': False, 'shown': None, 'total': None, 'truncated': False, 'coverage_from': None, 'retain_days': None, 'note': None}, 'impact': 12.6, 'status': 'resolved', 'first_seen': 1793677716, 'last_seen': 1794904001, 'resolved_at': 1794905754, 'acked_by': None, 'acked_at': None, 'snoozed_until': None, 'occurrences': 193, 'narrative': "studiosvc's p95 is 23.9 s on automation, above a normal of 489 ms ± 446 ms for this hour of the week (4 of the last 4 five-minute buckets, z = 52.4, 5 weeks of history). Whether users noticed depends on which endpoints carried it — the service page shows the split. A p95 that stays under a second is never reported — below that nobody notices, whatever the deviation.", 'narrative_kind': 'template', 'discriminator': 'api_p95_ms', 'level': 'low', 'told_level': None, 'episode_at': 1794896955, 'chat_thread': None, 'kind_label': 'Latency anomaly', 'group': 'Anomalies', 'age_s': 1228052, 'stale_s': 1767}
- 2026-11-17 09:22:05 UTCreopened x-ray · severity=info was=info title=studiosvc p95 9.50 s vs a normal 489 ms ± 446 ms for Tue 08:00 UTC (4 of the last 4 checks) level=low was_level=low finding={'id': 15924, 'fingerprint': '2cc6f9770ec51367864a', 'kind': 'anomaly.latency', 'severity': 'info', 'cluster': 'automation', 'subject': 'studiosvc', 'title': 'studiosvc p95 9.50 s vs a normal 489 ms ± 446 ms for Tue 08:00 UTC (4 of the last 4 checks)', 'facts': {'measure': 'api_p95_ms', 'unit': 'ms', 'value': 9500.00000000003, 'expected_median': 488.75, 'expected_mad': 446.0155, 'z': 20.203894259280293, 'rel': 18.437340153452745, 'buckets_hit': 4, 'n': 34, 'weeks': 5, 'slot': 'Tue 08:00 UTC', 'bucket': 1794905700, 'window': [1794904800, 1794906000], 'floor_ms': 1000, 'recent_deploys': [], 'discriminator': 'api_p95_ms'}, 'evidence': [{'label': 'Service', 'href': '/service/studiosvc?cluster=automation&from=1794883200&to=1794906000'}, {'label': 'Fleet vs normal', 'href': '/fleet?cluster=automation&from=1794883200&to=1794906000'}], 'provenance': {'source': 'rollup', 'collected_at': 1794906458, 'age_seconds': 881, 'stale': False, 'shown': None, 'total': None, 'truncated': False, 'coverage_from': None, 'retain_days': None, 'note': None}, 'impact': 7.8, 'status': 'open', 'first_seen': 1793677716, 'last_seen': 1794907325, 'resolved_at': None, 'acked_by': None, 'acked_at': None, 'snoozed_until': None, 'occurrences': 194, 'narrative': "studiosvc's p95 is 23.9 s on automation, above a normal of 489 ms ± 446 ms for this hour of the week (4 of the last 4 five-minute buckets, z = 52.4, 5 weeks of history). Whether users noticed depends on which endpoints carried it — the service page shows the split. A p95 that stays under a second is never reported — below that nobody notices, whatever the deviation.", 'narrative_kind': 'template', 'discriminator': 'api_p95_ms', 'level': 'low', 'told_level': None, 'episode_at': 1794907325, 'chat_thread': None, 'kind_label': 'Latency anomaly', 'group': 'Anomalies', 'age_s': 1229623, 'stale_s': 14}
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
| source | rollup |
| collected_at | 1794911375 |
| age_seconds | 682 |
| stale | False |
| shown | None |
| total | None |
| truncated | False |
| coverage_from | None |
| retain_days | None |
| note | None |
| detector | seasonal |