Buoyant Cloud

• Out-of-memory events now conclude automatically once a workload stops reporting OOM activity, resolving the corresponding notifications

• Buoyant Enterprise for Linkerd now receives an explicit message when a subscription is no longer active, rather than a generic authentication failure

• Enabled enterprise.buoyant.io access for all organizations
• Stale events are now cleaned up automatically

• Surfaced the Buoyant Enterprise for Linkerd license IP address for allowlisting

• Added Trust Anchor Rotation support: track rotation phase and convergence across clusters

• Added a status page link to the top navigation

• Reduced false positives from stale proxy limit events

• The installer script now includes the linkerd-trust CLI

• FIPS builds are now served for -fips version checks, and via FIPS=1 in the install script

• Reduced initial page load by moving the Topology renderer out of the main bundle

• Faster initial page load: routes are now lazy-loaded

• Added error boundaries, so a failure in one panel no longer blanks the page

• Fixed inconsistent proxy limit alerting

• Assorted table and layout fixes

• The install dialog now works with browser back and forward navigation
• Fixed markdown rendering of inline code and spacing

• Time range is now stored in the URL for Metrics
• Fixed broken Topology graph styles

• Buttons and links inside a clickable table row are now independently clickable

• Linkerd certificate events now account for webhook root CAs
• Added domain name tooltips to HTTPRoutes
• Table rows are now fully clickable

• Fixed sidecar container ports missing from workload details

• Microsoft Teams webhooks are now validated when added or updated

• Improved Microsoft Teams notification error reporting

• Fixed a race condition when filtering Events

• Simplified the Enterprise installation page
• Fixed drag-to-zoom on Metrics charts
• Success rate chart tooltips now sort consistently

• Improved license expiration warnings in the UI

• Added ability to visually select and zoom in on any given graph on the Metrics page

• Added custom time range selector on the Metrics page

• Added native sidecar support for Linkerd proxies

Updated Topology which includes the following improvments:

• Improved performance on larger meshes
• Simplified node label
• Added workload sidebar when node is clicked
• Moved node controls to sidebar
• Added workload metrics for latency and resource usage
• Added support for TLS type highlighting
• No longer disabling search bar when node is pinned

• Added default alerting rules for webhook certificates

• Fixed PagerDuty integration

• Added Datadog multi-region support

• Added Services table on Workload Detail page

• Improved clickthrough from Workload Detail page to top-level Metrics page

• Improvements to proxy diagnostic graphs

• Added support for HTTPRoutes from the Gateway API

• Added proxy diagnostic graphs to the Linkerd proxy detail page

• Added time series time window selector to Container detail page

• Added support for HTTPRoutes that reference Services

• Added Argo Rollouts support in the “Related events” table for Managed Linkerd events

• Updated Managed Linkerd to restart argo rollouts
• Update traffic data to include outbound TCP edges

• Added filter in Events list for Linkerd version alerts

• Added links to the top-level metrics and topology pages from the workload detail page

• Added integration to send Buoyant Cloud events to Microsoft Teams

• Added integration to send Buoyant Cloud events and metrics to Datadog

• Improvements to Topology search
• Added zoom controls in Topology
• When a node is pinned in Topology, the full filter string is honored, which may result in not showing all clients and dependencies

• Deprecated Buoyant Cloud agent v0.7.2 and older

• Added feature to send diagnostics to Buoyant for the control plane

• Added Managed Linkerd support for stable-2.12.4
• Fixed bug where manual rollouts were being attributed to Managed Linkerd

• Failed and canceled event status are now reflected in email and Slack notifications
• Label workloads as “unmeshed” if all running proxies are in pods that are not running

• Added a success rate column to the Traffic page
• Always show ports on Topology edges

• Added time window selectors to the Traffic and Topology pages

• Updated Topology page to use the same data source as the Traffic page

• Removed default alerting rule for control plane p95 latency

• Improved denied traffic and protocol detect alerts
• Updated policy columns on Traffic and Workloads pages

• Added severity to Slack event notifications
• Added a bar chart of Linkerd events to the Overview page
• Removed pods table on the events detail page when a rollout fails

• Added support for Linkerd 2.12 authorization policy on the workload detail policy tab

• Added Linkerd health and managed Linkerd status columns to the Linkerd deployments table on the Overview page

• Added detail pages for all containers

• Added the ability to filter events by alert type
• Updated the default threshold for control plane p95 latency alerts

• Added containers table on workload detail page
• Added default alerting rule that monitors percent of resource limit for proxies

• Removed all legacy control plane workload alerts, and added default control plane alerting rules

• Updated the overview page to be more Linkerd-focused