What's new this week

This section summarizes what was released this week.

Fast lane GA features

These features are generally available for all customers. Slow lane customers and tenants will have these features automatically enabled on December 9, 2026.

alerts: URL values in an alert's context are rendered as links. Engineers can open the runbook, dashboard, or ticket an alert points to in one click instead of copying the URL.
ingest: Ingest routing is on by default on the Canada, Asia-Pacific, Australia, and EU-2 clusters, together with the Ingest routing settings page and the source datasets explorer. Teams can see and manage where incoming data lands from the UI. Other clusters follow in a later wave.

Slow lane GA features

These features are enabled for all slow lane customers and tenants today. Fast-lane GA was 2026-10-14.

ai-sre: The Ask User Question panel is compact and can be minimized, keeping the chat viewport clear while waiting for a user decision during an investigation.
ai-sre: Oversized chat image uploads are downscaled instead of rejected, so screenshots and diagrams of any size can be shared with the investigation agent.
apm: Service Explorer differentiates Service, Database, and Message Broker node types across the catalog, health, alerts, and monitor APIs. The service catalog accepts a serviceType filter, and alerts resolve which kind of service triggered them.
input-picker: The redesigned input picker, with unified search across datasets, resources, and scopes.
monitors: Individual webhook notification headers can be marked as redacted. Marked headers read back as ***REDACTED***, keeping API keys and bearer tokens hidden from other users who view or test an action configuration.

Public preview and private preview features

features are available by invitation only. features are available by inquiry.

apm: External calls are recorded in a dedicated Tracing/External Call Metrics dataset, and each external peer is counted once, at the innermost span of the call, for selected customers.
datasets: Per-dataset freshness is emitted as dataset_freshness rows in the System datastream for selected customers, so freshness can be monitored and charted like any other data.
monitors: The monitor create, read, and update operations of the /v1/monitors REST API move from alpha to beta. Monitors created through the API can set which alert severities start an AI-SRE investigation and how many investigations a monitor may start per window, and the monitor statistics endpoint can count values of multi-value attributes.

Bug fixes

The following issues are fixed in this week's release:

ai-sre: A toast now explains when forking a chat fails, instead of failing silently.
apm: Correlation tags are applied on the exceptions view, and external nodes on the trace service map can be selected.
authorization: Permission-denied errors no longer include access key values. (OB-66384)
aws-explorer: Long table cell text is truncated with an ellipsis instead of overflowing the cell.
dashboards: Legacy dashboards that contain an unknown view type convert to the v2 definition instead of failing. (OB-72220)
dashboards: A filter with an error now shows its invalid state on the dashboard. (OB-71134)
datatable: JSON key columns keep their order and can be dragged to a new position.
documents: Downloading an uploaded document is restricted to workspace admins. (OB-72206)
export: Cursor-based query exports no longer loop endlessly.
monitors: Alert reminders resume when an alarm is extended.
opal: The OPAL filter bar applies text set from outside the editor immediately, and the autocomplete popup closes after a clause is typed and stays steady while arrowing through suggestions. (OB-71750)
support: Support requests send code and literal text to the support ticket exactly as typed. (OB-71975)
traces: The trace view explains span errors that are hidden by the current view, and span events fall back correctly when the primary source is missing. (OB-71678, OBSSD-5863)

APM Tracing Content Rollout

This week's tracing content changes are summarized from the APM Upgrade Impact — 2026-11-09 report, which has per-cluster details.

Non-breaking changes rolling out this week

Tracing/Service takes the last observed language and host type, so a service first seen without telemetry.sdk.language or k8s.cluster.uid picks them up once later spans carry them.
Error Tracking Metrics escapes regex literals in the exception message fingerprint, so exception grouping matches the span error path.
Service inspector and Service Explorer drilldown counts, duration sums, and distinct caller counts are scaled by canonical-trace sampling, so head-sampled traffic estimates whole-traffic volume. Duration t-digests stay unscaled.
ProduceExternalCalls and UseExternalCallMetricsDataset will be flipped on by default in the next scheduled breaking upgrade.

Breaking changes flipping this week

None — no feature flag flips affect a majority of in-window customers this cycle.

REST API changes

REST API surface changes for this release are published separately in the REST API Changelog, which covers version changes, breaking changes, deprecations, and migration.

Terraform provider

The following updates are available in the Observe Terraform provider.

v0.14.76 (2026-11-09)

New features

observe_monitor_v2: Add ai_triaging_mode and AI-SRE investigation limits. Alert-triggered AI-SRE investigations are available by invitation.

v0.14.75 (2026-11-06)

Bug fixes

Documentation links to Observe Regions use absolute URLs.
The v2 dashboard example gives query cards an id.

Coming soon

The following GA features are already available for fast lane customers, and will be enabled for slow lane customers and tenants next week.

ai-sre: General and Investigation mode runs, including alert-triggered investigations, can take up to 50 steps instead of 25.
ai-sre: The question panel slides up from under the chat composer instead of replacing it, so a draft is kept while answering or skipping a question.
dashboards: Dashboards can be exported as PDF.
service-explorer: The service inspector header has an Investigate service button that opens AI-SRE and starts an investigation of that service.