Jun 25 to July 1

Features and Enhancements

  • OpsGenie users can now get paged by Weave Cloud when their cluster is needs attention. OpsGenie is an alerting and incident management service that competes with PagerDuty. The Weaveworks team uses it for the Weave Cloud production environment.

  • Disabled “Update” buttons on read-only Deploy clusters now show a tooltip explaining why they are disabled.
  • If you are using Helm charts for your deployments, a command and instructions are now available for how to install the Weave Cloud agents.

  • If you are using the Flux Helm integration, you can now see your Helm workloads in Weave Cloud.

  • Read-only workloads are now indicated on its corresponding deployment information page.

Bug Fixes

  • When switching between instances, Weave Cloud now correctly updates the list of namespaces in the monitoring dashboards.
  • Weave Cloud no longer displays gratuitously inappropriate auto-complete suggestions in the PromQL aggregator clauses (i.e. the by bit in sum(metric_name) by (...).
  • When deploying a workload without containers, Weave Cloud displays a helpful error message instead of crashing.

Jun 18 to June 24

Features and Enhancements

  • Deployments and their history are now shown on the timeline. Mouseover the blue dots to view the details of the deployment.

  • The release plan displays default cluster resources monitoring dashboards, if you haven’t yet set up specific notebooks for the workload.

Bug Fixes

  • You can now add custom filter tags for your non-automated deployments.
  • In the past, Weave Cloud redirected you to the deployments Git configuration page, even though you already configured deploy to work with Git. This has now been fixed and no longer occurs.
  • Weave Cloud now creates the correct AWS CloudWatch configuration during the installation of the Weave Cloud agents.
  • Switching context from one area of Weave Cloud to the “Workloads” screen preserves the current time in timeline. This wasn’t always the case.
  • Empty charts appeared when you clicked an object in the legend of a chart on the Workloads dashboards and then switched namespaces. This has been fixed.
  • Selecting “Table view” in notebooks now shows that view more consistently.

Jun 11 to June 17

Features and Enhancements

  • URLs have been changed and now refer to their Weave Cloud feature names (e.g. “deploy”), rather than their open source project counterparts (e.g. “fluxv2”).
  • Instructions for both command line and web console are now available for AWS CloudWatch integration.

  • By default, email notification are sent for every event. All notifications link to the Notifications settings page from those emails where you can disable the notification if you need to.

Bug Fixes

  • Deployments now work with repositories that contain cron job definitions.
  • Adding a receiver address (e.g. an email) on the notifications settings page no longer unchecks all boxes checked by default.
  • Events can now be re-enabled after disabling all events from the notification settings screen.
  • Clicking on a workload now takes you to graphs that show only that workload, rather than whole-cluster graphs.
  • You can now select a time frame for AWS dashboards.
  • When waiting for repositories to be loaded in the deployments page, there is no longer an error message.
  • HTML code no longer shows in desktop notifications.
  • Mistyping an instance name in a Weave Cloud URL displays a 404 error page, rather than a blank page like it did before.
  • The correct percentages are now shown on GPU dials (if you are using GPU workloads).