v0.134.0

supercorp-ai/supergatewayv0.134.0Jul 22, 2026by github-actions[bot]

AI Summary

This release introduces major updates to LLM observability, dashboard v2, and infrastructure monitoring v2, including new GCP Memorystore Redis support and product analytics instrumentation. Users upgrading with a PostgreSQL metastore must apply a specific database migration workaround to prevent startup failures.

Key Highlights

Critical Migration Issue: Upgrading with a PostgreSQL metastore requires a manual SQL workaround to bypass a failing migration and prevent SigNoz from crashing.
LLM Observability: Introduces AI observability dashboards, side navigation, and model pricing overviews.
Dashboard V2: Enhances variable runtime, allows panel export, and adds alert condition pre-population.
Infrastructure Monitoring V2: Adds pod metrics, status counts, onboarding checks, and chart documentation.
GCP Memorystore Redis: Adds support for monitoring Google Cloud Memorystore for Redis services.

New Features

GCP Memorystore Redis service integration
Span mapper test endpoint
Product analytics instrumentation for list and detail pages
Dashboard link sharing with current variable selection
Telemetry resource provisioning support
LLM observability dashboard and side navigation
Quick filters support for field values
Table export functionality for Logs and Traces
Infrastructure Monitoring V2: Pod metrics, counts cards, and onboarding checks
Dashboard V2: Variable runtime and panel export

Full Release Notes

## ⚠️ Upgrading to v0.134.0 with a PostgreSQL metastore

A database migration in this release fails on PostgreSQL, so SigNoz does not come up after the upgrade and keeps restarting. Fresh installations and installations with a SQLite metastore are not affected. Please check out #12293 for more details.

As a workaround, mark migration 101 as done in the migration table and restart SigNoz:
```sql
# Replace <last_group_id> with the group_id of the last applied migration in the migration table. 
INSERT INTO migration (id, name, group_id, migrated_at)
VALUES (101, 101, <last_group_id> + 1, '2026-07-27 14:24:16');


```
## 🚀 Features

* feat: adding gcp memorystore redis service (#12118) @swapnil-signoz
* feat: span mapper test endpoint (#11795) @nityanandagohain
* feat(dashboard-v2): product analytics instrumentation for list + detail pages (#12199) @AshwinBhatkal
* feat(infra-monitoring): persist custom time on drawer across categories (#12038) @H4ad
* feat(dashboard-v2): share a dashboard link with the current variable selection (#12198) @AshwinBhatkal
* feat(authz): add support for telemetry resource provisioning (#12168) @vikrantgupta25
* feat: rename llm o11y to ai o11y [2/2] (#12141) @tewarig
* feat: add llm observability to side navigation [1/3] (#12123) @tewarig
* feat(authz): enable FGA for telemetry resources (#12095) @vikrantgupta25
* feat(quick-filters): add support for /v1/fields/values (#11788) @H4ad
* feat: table export in Logs and Traces explorer table tabs (4) (#12071) @aks07
* feat(dashboards-v2): pre-populate alert condition from panel Reduce To + thresholds (#5291) (#12137) @ahrefabhi
* feat: add dashboard in llm o11y (#12136) @tewarig
* feat(llm-attribute-mapping): group + mapper drawers + diff-save [3-4/5] (#11780) @tewarig
* feat(dashboard-v2): variable runtime — resolver, fetch engine, panel loading, impact dialog (#12125) @AshwinBhatkal
* feat(dashboard-v2): export panels & dashboards from explorers (#12043) @ahrefabhi
* feat(infrastructure-monitoring-v2): pod status/counts & couple more fixes (#12127) @H4ad
* feat(infrastructure-monitoring): add pod metrics tab (#12106) @H4ad
* feat(infrastructure-monitoring-v2): add onboarding checks (#12119) @H4ad
* feat(infrastructure-monitoring-v2): add counts cards (#12120) @H4ad
* feat(infra-monitoring): add docs for every chart (#12037) @H4ad
* feat(infrastructure-monitoring-v2): pod status/counts & couple more fixes (#12127) @H4ad

## 🐛 Bug Fixes

* fix: add rank column to tag relations to keep note of tag ordering in dashboards (#12170) @namanv3
* fix: remove omitempty/omitzero and explicit empty value handling in dashboard v2 spec (#12197) @namanv3
* fix(alerts-v2): preserve loaded thresholds on edit-page refresh (#12226) @ahrefabhi
* fix: fetch latest monotonicty for metrics instead of any random (#12205) @namanv3
* fix: show the backend error message in the error modal instead of a generic axios error (#12215) @AshwinBhatkal
* fix(dashboard-v2): validate key:value tags in the editor (inline-edit + backend rules) (#12062) @AshwinBhatkal
* fix(overlayscrollbars): leaking memory due to img.load event listener (#12179) @H4ad
* fix(session): validate ref and callback state against global allowed_origins (#12172) @vikrantgupta25
* fix(dashboards-v2): panel editor Save/dirty, refresh retention, dropdown clipping & time-window fixes (#12146) @ahrefabhi
* fix(onboarding): surface Temporal Cloud Metrics as a metrics integration (#12203) @crazyuploader
* fix(uplot): missing destroy uplot on unmount (#12182) @H4ad
* fix(metrics): resolve any filter key context instead of erroring (#12188) @tushar-signoz
* fix(clickhouseprometheus): resolve __name__ matchers against metric_name and anchor matcher regexes (#12154) @srikanthccv
* fix(querybuilder): resolve data-type collisions for numeric intrinsic columns (#12176) @tushar-signoz
* fix(querybuildertypesv5): round-trip unset stepInterval and empty field-key values (#12171) @therealpandey
* fix(query-search): do not override token classes (#12139) @H4ad
* fix(logs-explorer): prevent long json body from overflowing (#12173) @aks07
* fix: css fixes for llm o11y model pricing and overview  (#12145) @tewarig
* fix(dashboards-v2): round-trip all zero-valued spec fields (#12162) @therealpandey
* fix(querybuildertypesv5): allow empty source, temporality and timeAggregation in the metric query spec (#12164) @therealpandey
* fix(clickhouseprometheus): include hour-floored registration rows at the query window end (#12153) @srikanthccv
* fix: convert key not found to warnings for logs (#12134) @tushar-signoz
* fix(clickhouseprometheus): merge fingerprints sharing a labelset instead of injecting a fingerprint label (#12152) @srikanthccv
* fix(querybuildertypesv5): keep subnormal values finite in roundToNonZeroDecimals (#12157) @srikanthccv
* fix(dashboards-v2): round-trip zero-valued threshold + query fields (#12158) @therealpandey
* fix: has function family revamp (#12089) @tushar-signoz
* fix(querybuildertypesv5): guard float overflow in roundToNonZeroDecimals (#12151) @srikanthccv
* fix(ruletypes): round-trip empty notificationSettings fields (#12138) @therealpandey
* fix(infra-monitoring): use proper metric names corresponding to entities to be counted (#12130) @nikhilmantri0902
* fix: handling missing keys metadata (#12091) @tushar-signoz
* fix(onboarding): surface Temporal Cloud Metrics as a metrics integration (#12203) @crazyuploader
* fix: css fixes for llm o11y model pricing and overview  (#12145) @tewarig

## 🧰 Maintenance

* chore: add integration tests for metrics under reduction - query part (#11979) @srikanthccv
* chore: pass org id for flagger (#12126) @tushar-signoz
* refactor: gcp integration memorystore redis dashboard update (#12229) @swapnil-signoz
* test(e2e): cover alert edit-page threshold persistence (#12227) @ahrefabhi
* refactor(dashboards-v2): dedupe threshold color palette into an enum & reuse normalizeOperator for alert prefill (#12142) @ahrefabhi
* refactor(dashboard-v2): organize variable subsystem into components/ hooks/ utils/ (#12135) @AshwinBhatkal
* refactor(infrastructure-monitoring-v2): ui/ux improvements (#12191) @H4ad
* test(e2e): fix flaky dashboards-detail specs (sidenav overlay + slow hydration) (#12180) @AshwinBhatkal
* test(e2e): fix teardown crashes (network endpoint race + eager keeper lookup) (#12204) @AshwinBhatkal
* chore: pass org id for flagger (#12126) @tushar-signoz

## New Contributors
* @crazyuploader made their first contribution in https://github.com/SigNoz/signoz/pull/12203

**Full Changelog**: https://github.com/SigNoz/signoz/compare/v0.133.0...v0.134.0