Skip to content

Latest commit

 

History

History
499 lines (308 loc) · 11.6 KB

CHANGELOG.md

File metadata and controls

499 lines (308 loc) · 11.6 KB

Changelog

v0.120

Changed

  • Karma no longer uses alertmanager api/v2/status to discover cluster peers. It will only rely on cluster field configuration options set in karma config file.

v0.119

Fixed

  • silenceForm.strip.labels did't strip labels that are in the filter bar #5817.

v0.118

Fixed

  • Fixed listen:prefix handling that was broken in v0.117 - #5623.

v0.117

Changed

  • Migrated project from Create React App to Vite.

v0.116

Added

  • Added headers option to history:rewrite config block. This allows to set custom headers passed to Prometheus when sending history query requests.

v0.115

Fixed

  • Improved memory usage.

v0.114

Fixed

  • Fixed silcen form crashes when silenceForm:defaultAlertmanagers config option is not set.

v0.113

Added

  • Added silenceForm:defaultAlertmanagers config option - #5086 (@david-caro).

v0.112

Fixed

  • Upgraded project dependencies.

v0.111

Fixed

  • Fixed duplicated @receiver labels showing both on the alert and in the footer.

v0.110

Fixed

  • Alert group footer was always hidden when displaying only one alert, even if there were shared labels or annotations to display - #4892.

v0.109

Added

  • Added ability to delete multiple silences from the silence browser modal - #4618.
  • Added build_info metric - #4764.

Fixed

  • Managing silences for alertmanager instances with / in the name was failing. This release adds a workaround for it - #4674.

v0.108

Fixed

  • Fixed CSS glitches on Google Chrome 105.

v0.107

Changed

  • Don't show fatal error page if configured health checks are failing but alertmanager is still showing alerts. A popup message with will still be visible.

v0.106

Added

  • proxy_url option for history rewrite rules #4510 (@tolleiv).

Changed

v0.105

Added

  • Allow to copy alert payload from alerts dropdown menu #4378.
  • Clicking on a label while holding Shift will now copy its value to clipboard #4378.

Changed

  • Change max value of alertsPerGroup to 25 #4420.

v0.104

Added

  • Added @inhibited and @inhibited_by filters #4397.

Changed

  • Dropped support for alertmanager older than 0.22.0.

v0.103

Fixed

  • Fixed a regression in alert history queries.

v0.102

Fixed

  • Correctly set filter history.
  • Correctly escape label values when quering Prometheus for alert history.

v0.101

Fixed

  • Don't crash on autocomplete errors.

v0.100

Changed

  • silences:expired option no longer takes alerts age into account.
  • Moved silence progress bar to the bottom of the silence block.
  • Removed sentry support to reduce application size. sentry:private and sentry:public config options are no longer valid.

v0.99

Fixed

  • Fixed regexp escaping when editing silences #3936.

Changed

  • Don't render @cluster labels if there's only one cluster configured #3994.
  • Show ! on favicon badge if there's any alertmanager upstream with errors #3987.

v0.98

Fixed

  • Fixed regexp escaping for auto-populated silence matchers #3936.

Changed

v0.97

Fixed

  • Fix escaping regex values in when editing silences #3936.
  • Fix UI handling of 401 errors #3942.

v0.96

Fixed

  • Fix escaping regex values in silence form #3936.

v0.95

Fixed

  • Messages are now logged correctly when both --log.format=json and --log.timestamp=true flags are set #3822.
  • Escape label values in silence form #3866.
  • Some silenced were showing incorrect alerts count #3909.

Changed

  • Show a placeholder message if no alertmanager server is configured, instead of failing to start.

v0.94

Added

  • tls options to history:rewrite rules, allowing customising TLS options for requests made by karma to Prometheus servers when querying alert history, #3707.
  • Unsilenced alerts will now show recently expired silences if they are old enough. By default silences expired in the last 10 minutes will be shown, this can be configured by setting silences:expired option or --silences.expired flag. Setting this value to 5m will show silences if they expired in the last 5 minutes but only if the alert started firing at least 5 minutes ago.
  • alertAcknowledgement:comment will replace %NOWLOC% string with a timestamp formatted using local time zone, use %NOW% for timestamps using UTC timezone, #3704.

v0.93

Fixed

  • Toast popup messages didn't respond to clicks.
  • Alertmanager with brackets in the name wasn't able to create, edit or delete silences #3651.

Added

  • Loading user groups from HTTP headers, #3361 (@supertassu).
  • Added labels:keep_re, labels:strip_re and labels:valueOnly_re config options #3659 (@aalexk).
  • labels:order config option to allow customising order of labels #3500.

Changed

  • Refactored internal APIs.
  • Overview modal won't show label name for every value to save screen space.
  • Retry failed requests when collecting alerts and silences from alertmanager.

v0.92

Fixed

  • Overview modal could show incorrect counters for @state labels.

v0.91

Fixed

  • Alert history rewrite rule wouldn't work unless they had a / suffix in source field, this is now fixed and rewrite rules works as documented.
  • Clicking Silence this group could generate duplicated label matchers #3509.

v0.90

Fixed

  • Alert history will now correctly handle Prometheus servers with --web.external-url containing sub-uri, see #3387.
  • Alert history queries didn't include grid labels as those got removed from alert groups in #3222, this is now fixed.

Changed

  • Reduced CSS bundle size by removing unused CSS rules with PurgeCSS.
  • Upgraded bootstrap to 5.1

v0.89

Fixed

  • API could error if multiple alertmanager upstreams belonging to the same cluster were configured, but cluster option was unset #3372 (@valihanov).

v0.88

Added

  • Added a dedicated API endpoint for silence previews.
  • Added a dedicated API endpoint for overview modal.
  • Individual alert details are now lazy-loaded to improve performance when dealing with a huge number of alerts per group.
  • Added /version endpoint returning karma and Go runtime version #3332.
  • Added labels:valueOnly config option, see #3221.

Changed

  • Refactored internal APIs.
  • Grid labels are no longer shown on alert groups #3222. To ensure that those labrls are always visible swimlanes are now using position: sticky.

v0.87

Added

  • Use uber-go/automaxprocs to automatically adjust GOMAXPROCS to match Linux container CPU quota. Runtime value of GOMAXPROCS is now exported as a go_max_procs metric.
  • API will now return only first 40 alert groups by default, the rest can be loaded when user clicks on the Load more button. grid:groupLimit config option was added to customise how many groups are returned and displayed in the UI by default.

Changed

  • Automatic grid label selection logic was tweaked to avoid splitting alert groups.

v0.86

Added

  • Added support for alertmanager v0.22.0 negative matchers when creating/editing silences and in ACL rules.
  • Silence ACL rules no longer default isRegex to be false for filters and matchers, see ACL rules for updated docs.

Changed

v0.85

Fixed

  • History queries were always failing due to wrong Prometheus API usage.
  • URI handling for silence requests when proxy is used #3060.

Added

  • Ability to rewrite source URIs for alert history via history:rewrite config section #3064.

v0.84

Added

  • Added grid:auto config section for fine tuning automatic label selection for multi-grid, when multi-grid is configured to Automatic selection in the UI or when ui:multiGridLabel config section is set to @auto.
  • Added alert history estimation, see README.md and docs/CONFIGURATION.md for details.

Changed

  • karma will no longer fail to start if config file contains multiple alertmanager instances with different name but identical URI #3024.
  • Minor UI tweaks.

v0.83

Added

  • Added a new option on the list of labels used for multi-grid. When set in the UI to Automatic selection selection or @auto in ui:multiGridLabel config option karma will try to select the best grid label based on current alerts.

v0.82

Fixed

  • Header values are now sanitised before logging when log:config is enabled #2930.
  • Fixed a deadlock issue that could cause karma to hang #2944 (@jonaz).

Added

  • listen:timeout:read and listen:timeout:write config options for setting HTTP server request read and response write timeouts.
  • annotations:enableInsecureHTML config option #2886.

v0.81

Fixed

  • Fixed a deadlock issue that could cause karma to hang #2888.

Added

  • annotations:actions option allowing to move some annotations to alert dropdown menu #2596.
  • proxy_url alertmanager option to use when a proxy sever needs to be used for requests from karma to alertmanager API #2903 (@mhrabovcin).

Changed

  • Multi-grid label dropdown will only show label names from visible alerts.

v0.80

Added

  • Add /robots.txt to block search engine crawlers.
  • Easily change multi-grid source label via quick access dropdown on the grid header.

Changed

  • Reworked how notifications are displayed.

v0.79

Fixed

  • Not all labels were stripped when using lables:keep or labels:strip option #2585.

Added

  • healthcheck:visible alertmanager option to control if healtcheck alerts should be visible in the UI #2614.

v0.78

Fixed

  • Don't reset regex toggle when adding new silence labels #2520

Added

  • Added support for DeadMansSwitch alerts via healtcheck:alerts option on alertmanager upstream configuration #2512. Example:

    • Setup always on alert in each Prometheus server (prom1 and prom2):

      - alert: DeadMansSwitch
        expr: vector(1)
    • Add healtcheck configuration to karma:

      alertmanager:
        servers:
          - name: am
            uri: https://alertmanager.example.com
            healthcheck:
              filters:
                prom1:
                  - alertname=DeadMansSwitch
                  - instance=prom1
                prom2:
                  - alertname=DeadMansSwitch
                  - instance=prom2

    If any of these alerts is missing from alertmanager karma will show a warning in the UI.

v0.77

Fixed

  • Docker images reported version as dev #2479.

Changed

  • Alert groups will be rendered with fewer details when idle.

v0.76

Fixed

  • Fixed release builds using Github Actions

v0.75

Fixed

  • Fixed auth bypass for /health and /metrics endpoints. Those endpoints should be always excluded from authentication but that was broken in v0.73 #2465.

Added

  • listen:tls:cert and listen:tls:key config options for listening on HTTPS port

Changed

  • ghcr.io/prymitive/karma is now used as the primary repository for docker images instead of Docker Hub
  • alertAcknowledgement:commentPrefix config option was replaced by alertAcknowledgement:comment that can be used to customise the entire comment.