Releases: robusta-dev/robusta
0.14.0
Major Announcements
HolmesGPT Integration
Quickly find the root cause of Prometheus alerts, with the new HolmesGPT integration.
HolmesGPT is our most powerful AI-capability yet, and the first AI agent to achieve human-level accuracy when investigating incidents.
It uses an agent-based architecture to investigate problems the way that humans do - by iteratively gathering information until the picture is clear.
For more details, check out the HolmesGPT Integration Docs or watch a video of HolmesGPT in action.
Security Update
We upgraded Robusta Runner to 3.11 and removed many CVEs.
Breaking Changes
- All custom actions need to be Python 3.11 compatible.
- The robusta-cli pypi repo is being deprecated and replaced with robusta-api and any custom actions need to be updated accordingly.
Repository Updates
- Default image location is now Dockerhub (still available on GCR as well).
- Robusta CLI is now a separate project
What's Changed
- Holmes Integration by @arikalon1
- Updated Disk Benchmark Image by @arikalon1
- Robusta API Project by @Sheeproid
- Removed Robusta CLI by @Sheeproid
- Usage of Records Definition Instead of Records by @wrbbz
- Change Robusta Repo to Dockerhub by @Avi-Robusta
- Upgraded Robusta to 3.11 and Removed CVEs by @Avi-Robusta
- Popey Scans - ignore throttling warning by @RobertSzefler
- Custom warning events notifications by @Avi-Robusta
New Contributors
- @wrbbz made their first contribution.
Full Changelog: 0.13.1...0.14.0
0.14.0-alpha-1
Fix bug - wrong condition with HOLMES_ENABLED env var (#1472) * Fix bug - wrong condition with HOLMES_ENABLED env var * bump krr image
0.14.0-alpha
Major Announcements
HolmesGPT Integration
Quickly find the root cause of Prometheus alerts, with the new HolmesGPT integration. HolmesGPT is our most powerful AI-capability yet, and the first AI agent to achieve human-level accuracy when investigating incidents. It uses an agent-based architecture to investigate problems the way that humans do - by iteratively gathering information until the picture is clear. For more details, check out the HolmesGPT Integration Docs or watch a video of HolmesGPT in action.
[Add screenshots here]
Security Update
We upgraded Robusta Runner to 3.11 and removed many CVEs.
Breaking Changes
- All custom actions need to be Python 3.11 compatible.
- The robusta-cli pypi repo is being deprecated and replaced with robusta-api and any custom actions need to be updated accordingly.
Repository Updates
- Our images are moved to Dockerhub, and we will be phasing out GCP eventually.
- We have removed the Robusta CLI from this project to its own project
What's Changed
- Updated Disk Benchmark Image by @arikalon1
- Robusta API Project by @Sheeproid
- Removed Robusta CLI by @Sheeproid
- Usage of Records Definition Instead of Records by @wrbbz
- Change Robusta Repo to Dockerhub by @Avi-Robusta
- Upgraded Robusta to 3.11 and Removed CVEs by @Avi-Robusta
- Fixed Issue in Popey Job Scan Results by @RobertSzefler
- Improved Warning Events in the Robusta UI by @Avi-Robusta
- Added Holmes Integration to the UI by @arikalon1
New Contributors
- @wrbbz made their first contribution.
Full Changelog: 0.13.1...0.14.0
0.13.2-alpha
What's Changed
- update disk benchmark image by @arikalon1 in #1439
- Update docs on argocd install by @aantn in #1447
- Updated supported timezones and days by @pavangudiwada in #1444
- Robusta api project by @Sheeproid in #1449
- Update docs on deploying with argocd by @aantn in #1451
- Remove robusta cli by @Sheeproid in #1450
Full Changelog: 0.13.1...0.13.2-alpha
0.13.1
What's Changed
- bug fix - Fix auto recover mechanism for JWT token by @arikalon1 in #1445
- Fix race condition between job and pod creation when running scans by @arikalon1 in #1446
Full Changelog: 0.13.0...0.13.1
0.13.0
New Features
Quick Links
Define links to external systems
|
Define a new quick link |
Timeline Label Filters
Filter alerts by labels, across all clusters
What's Changed
- support annotations on template_enricher (#1422) by @zmx in #1423
- Evicted pod trigger by @Avi-Robusta in #1419
- Fix/zulip block formatting by @oscgu in #1404
- Send messages to slack via webhook by @romga in #1430
- Slack webhook prometheus labels by @romga in #1433
- New KRR version by @Avi-Robusta in #1437
- Support for custom service accounts by @Avi-Robusta in #1438
- Fix sinks activity intervals by @RobertSzefler in #1440
- increase krr default memory request to 2Gi by @arikalon1 in #1441
And much more
New Contributors
Full Changelog: 0.12.0...0.13.0
0.13.0-alpha-1
call execute on supabase rpc (#1443)
0.12.0
New Features
Slack Alert Grouping and Summarization Using Slack Threads
Reduce alert fatigue by grouping similar alerts and summarize them into a Slack thread.
Fully customizable based on severity, type of alerts, labels and more.
- slack_sink:
# other slack sink params
grouping:
group_by:
- cluster
interval: 86400
notification_mode:
summary:
threaded: true
by:
- identifier
- severity
New Zulip Sink
Receive Robusta alerts in Zulip. Contributed by community member @oscgu. See here for detailed instructions.
Other changes
- New Kubewatch version by @Avi-Robusta in #1395
- Added
nameOverride
andfullnameOverride
to helm chart by @kristeey in #1388 - Added
namespace_labels
support to the sink scope mechanism by @RobertSzefler in #1390 - Reduce alert noise with alert deduplication by @RobertSzefler in #1396
New Contributors
Full Changelog: 0.11.0...0.12.0
0.11.1-alpha
New Features
Slack Alert Grouping and Summarization Using Slack Threads
Reduce alert fatigue by grouping similar alerts and summarize them into a Slack thread.
Fully customizable based on severity, type of alerts, labels and more.
- slack_sink:
# other slack sink params
grouping:
group_by:
- cluster
interval: 86400
notification_mode:
summary:
threaded: true
by:
- identifier
- severity
New Zulip Sink
Receive Robusta alerts in Zulip. Contributed by community member @oscgu. See here for detailed instructions.
Other changes
- New Kubewatch version by @Avi-Robusta in #1395
- Added
nameOverride
andfullnameOverride
to helm chart by @kristeey in #1388 - Added
namespace_labels
support to the sink scope mechanism by @RobertSzefler in #1390 - Reduce alert noise with alert deduplication by @RobertSzefler in #1396
New Contributors
Full Changelog: 0.11.0...0.11.1-alpha
0.11.0
New Features
Change Tracking Triggers
Trigger a playbook when a resource field changes. Learn more
customPlaybooks:
- name: "NotifyOnImageChange"
triggers:
- on_deployment_update: {}
change_filters:
include:
- image
actions:
- resource_babysitter: {}
Fine Grained Scope For Triggers
Triggers now have an extensive include/exclude definition. Explore examples
customPlaybooks:
- name: "PostgresWarning"
triggers:
- on_event_create:
scope:
include:
- attributes:
- "type=Warning, involvedObject.name=postgres, event.reason=FailedScheduling"
actions:
- create_finding: #
title: "Failed scheduling postgres"
aggregation_key: "FailedScheduling"
View Argo Rollouts in the Robusta UI
Users can now monitor and manage Argo rollouts directly within the Robusta UI. See PR
Other Changes
- Update dependencies to mitigate vulnerabilities by @arikalon1 in #1370 and @Sheeproid in #1372.
- Added KRR scan support for
Mattermost
sink by @IdeoG in #1374.
Breaking Changes
- Sink Matcher Identifier Update:
Some playbooks notifications identifiers have changed. (For example:image_pull_backoff_reporter
is nowImagePullBackoff
)
If you're usingSink Matchers
based on theidentifier
you will need to update your sinks configuration.
For example:
Image_pull_backoff_reporter
→ImagePullBackoff
job_failure
→JobFailure
krr_report
→KrrReport
pod_oom_killer_enricher
→PodOOMKilled
report_crash_loop
→CrashLoopBackoff
Full list of changed identifiers
job_failure -> JobFailure
image_pull_backoff_reporter -> ImagePullBackoff
krr_report -> KrrReport
pod_oom_killer_enricher -> PodOOMKilled
report_crash_loop -> CrashLoopBackoff
show_stackoverflow_search -> ShowStackoverflowSearch
argo_app_sync -> ArgoAppSync
scale_hpa_callback -> ScaleHpaAction
alert_on_hpa_reached_limit -> HpaReachedMaximum
daemonset_fix_config -> DaemonsetFixConfig
daemonset_silence_false_alarm -> DaemonsetSilenceFalseAlarm
report_rendering_task -> GrafanaReport
disk_benchmark -> DiskBenchmark
report_image_changes -> ReportImageChanges
http_get -> HttpGet
http_post -> http_post
http_put -> HttpPut
java_process_inspector -> JavaProcessInspector
pod_jmap_pid -> PodJmapPid
pod_jstack_pid -> PodJstackPid
job_restart_on_oomkilled_community -> JobRestartOnOomkilledCommunity
node_not_ready -> NodeNotReady
count_pod_creations -> CountPodCreations
volume_analysis -> VolumeAnalysis
python_profiler -> PythonProfiler
pod_processes -> PodProcesses
python_memory_allocations -> PythonMemoryAllocations
debugger_stack_trace -> DebuggerStackTrace
python_process_inspector -> PythonProcessInspector
python_debugger -> PythonDebugger
popeye_report -> PopeyeReport
volume_snapshot_error -> VolumeSnapshotError
volume_snapshot -> VolumeSnapshot
restart_loop_reporter -> CrashLoopBackoff
http_stress_test -> HttpStressTest
Generic finding key -> GenericFindingKey
Generic Change -> GenericChange
General scheduled task -> GeneralScheduledTask
crash_loop -> CrashLoop
- Resource Babysitter action Breaking change
Thefields_to_monitor
was removed fromresource_babysitter
action. If you added a custom playbook, with this action, you may need to update the action configuration, see here about how to be alerted on custom configuration changes
New Contributors
Full Changelog: Compare Versions