Backlog crush

Weekly Update

This week, we’ve been doing a ‘backlog crush’ where we work through a long list of bug fixes, bits of polish and small feature requests.

We’ve spent a lot of time working on alert noise (making sure that our systems aren’t erroring when they don’t need to be), but amongst that we also made time for plenty of product work. Here are a few highlights (with a full list at the bottom of the page, as usual).

Jira integration error logs

It’s easy to make a change in Jira that leads to your automations breaking (e.g. adding a new required field). To help understand and fix these issues, you can now find the latest error activity in the Jira Cloud integration page under Settings.

You can now filter the homepage by incident type

You can now filter your active incidents list to include incidents of a specific type. This means that if you’re running a morning standup on the Data team, you can quickly see if you’ve got any active data incidents.

Linking incidents in Slack

We recently added the ability to link incidents together from the dashboard. Now, you can do the same via Slack! Try /inc related to give it a go.

More rich text support in Slack

When posting messages to Slack, we now support nested bullet points and custom emojis, so you can make your messages look even more 🧑‍🍳💋.

🚀What else we’ve shipped

  • 🆕 In our API, you can use an incident's number (the 123 in INC-123) instead of its full ID
  • 🆕 Upgrade to the latest OpenAI models for our AI features
  • 🆕 We now show a loader in the sidebar whilst a postmortem is being exported
  • 🆕 We will backfill your alerts if you configure a new attribute
  • 🆕 You can now access test and tutorial incidents when listing incidents in the public API, by using the mode filter
  • 🆕 You can now customise the "Report a problem" button on status pages
  • 🆕 You can now detach incident attachments straight within slack
  • 🆕 You can now filter by the incident source (e.g. user / API) when creating announcement rules
  • 👷‍♂️ Improve error handling when exporting to Notion
  • 👷‍♂️ Improve our Sentry integration to ensure we handle API errors gracefully
  • 👷‍♂️ Increase the cadence at which we check our integration connections, to spot errors faster
  • 👷‍♂️ Made our Salesforce, PagerDuty and GitHub integrations more robust, ensuring we notify you if it becomes disconnected
  • 👷‍♂️ We now parse Jira errors in a better way when syncing an incident ticket
  • 💅 Alert routes and sources no longer show as clickable to users without permissions
  • 💅 Display long status names nicely in the incident progress bar
  • 💅 For users that have multiple organisations, the dashboard link in #incidents announcement channels now will direct you to the desired organisation's dashboard.
  • 💅 Improve documentation on how automatic GitHub linking works
  • 💅 Improve handling of child/parent relationships of Jira tickets
  • 💅 Improve icon alignment and size in follow-up overflow menu
  • 💅 Improve layout of sections in alerts details page
  • 💅 Improve spacing in incident lifecycle conditions
  • 💅 Improve the loading experience when syncing large sets of data in both PagerDuty and Salesforce integrations
  • 💅 In the dashboard, there was a 1 pixel gap between the border below our logo and the border below the page title. They now line up nicely.
  • 💅 Increase the number of workflow steps that can be run to 200
  • 💅 Make our announcement thread deletion (when cancelling/declining an incident) more resilient to race conditions.
  • 💅 Options to assigning roles and tasks in the post-incident flow are now clearer
  • 💅 Pin the triage message (where you can accept / decline an incident) to the incident channel so it's easy to quickly find
  • 💅 Removed a call-to-action modal in the alert source form
  • 💅 Renamed Clickup to ClickUp in the catalog
  • 💅 Surface which post-incident flow statuses are being used when attempting to delete a post-incident flow
  • 💅 The /incident test form now points out that workflows can be configured to run on test incidents.
  • 💅 Threaded incident updates on announcement posts will now say "Incident resolved" rather than "Incident updated" if the incident update is transitioning out of an active incident
  • 💅 Use consistent margins for inputs that are required or optional
  • 💅 Make our dashboard more resilient to Slack's API being slow
  • 💅 We made the instructions on the "would you like to enter the post-incident flow?" modal more intuitive
  • 💅 We no longer trigger the post incident flow for tutorial incidents
  • 💅 We now show the type of each recent incident in the "is this part of a recent incident?" modal
  • 💅 We'll now stop you going 'back in time' on a statuspage beyond where we have incident data
  • 💅 When creating an internal status page, if you're trying to create over 200 components we'll show you how many are being hidden.
  • 💅 When you have a broken reference in the catalog, it'll now be easier to see that it's broken when editing a catalog entry as it'll have a red badge
  • 🐛 Correctly split a long summary to a threaded Slack message, when linking incidents
  • 🐛 Fix exporting a multi-select field to a Jira text field in a follow-up
  • 🐛 Fixed a bug where custom severity placeholders weren't shown in the web
  • 🐛 Fixed a bug where derived custom fields wouldn't always update when editing an incident's custom fields via the public API
  • 🐛 Fixed a bug where expressions weren't correctly being created in alert routing
  • 🐛 Fixed a bug where our variable badges would sometimes end up 'floating' on the page where they shouldn't be
  • 🐛 Fixed a bug where selecting a catalog attribute for a customer status page was disabled
  • 🐛 Fixed a bug where the templated text editor wasn't always correctly displaying expression variables
  • 🐛 Fixed the behaviour of the "Save policy" popup, to make it clear what it will do on save
  • 🐛 If an incident update only changes custom fields, we won't post that update to the channel.
  • 🐛 Return debrief placeholders in a consistent order
  • 🐛 Status page incidents with very long names were running off the side of the screen in the dashboard. They're now nicely truncated.
  • 🐛 Use a consistent definition of 'active' incident in the homepage
  • 🐛 We correctly escape Slack special characters in messages for alerts
  • 🐛 We fixed a bug where we were unable to generate a postmortem if a link didn't contain a URL
  • 🐛 We fixed a bug where we were unable to post updates to Slack when StatusPage incidents had long names
  • 🐛 We had recently started including the 'incident created' message we post at the start of an incident in the timeline. This was not helpful and has been removed.
  • 🐛 We now correctly set the default incident state of active or triage
  • 🐛 We've fixed the duplicate "Welcome to" message when adding a new announcement rule to the main incidents channel
  • 🐛 When editing an alert route, changing the setting to auto-resolve incidents would not enable the 'save changes' button. Now it does!
  • 🐛 You won't be able to add workflow conditions that are empty anymore

Operational excellence starts here