Releases: kyma-project/telemetry-manager
Releases · kyma-project/telemetry-manager
1.17.0
Release Notes
-
Promoted to stable: Telemetry Manager now checks the health of the data flow in agents and gateways, and reports it to the pipeline resources under a new status condition TelemetryFlowHealthy. With this, you can easily detect if there are problems with the data flow for logs, traces, and metrics; for example, caused by unavailability of backends (#425).
-
For better usability and consistency, the reasons and messages for status conditions have been extended and aligned (#1023)
-
Metrics: Metrics provided using a MetricPipeline input now have a well-defined OTEL instrumentationScope version attribute defined to identify the Telemetry module's version as origin of the metric (#1000).
-
Metrics/Traces: Update to OTel Collector to version 0.100.0 (#1085)
Changelog
Features
- 9197134: feat: Add Flow Healthy print column (#1138) (@skhalash)
- 347f42a: feat: Bump Fluent Bit to 3.0.4 (#1108) (@skhalash)
- 22abd6e: feat: Enable self monitor (#1136) (@skhalash)
- 90c8551: feat: Improve self-monitor troubleshooting guides (#1142) (@skhalash)
- e8de168: feat: Introduce crd categories (#1009) (@k15r)
- 0f4ede3: feat: Performance test and optimize self-monitor (#1111) (@hisarbalik)
- 924a25e: feat: Set module version in instrumentation scope (#1134) (@chrkl)
- d02ec4b: feat: Update to otel-collector 0.100.0 image version (#1085) (@TeodorSAP)
Bug fixes
- 6841696: fix: Cover all status False ConfigurationGenerated condition reasons within legacy conditions (#1119) (@TeodorSAP)
- b6980f7: fix: Set TelemetryFlowHealthy condition to False if the pipeline is non-reconcilable (#1144) (@shorim)
- e73ad8c: fix: Stabilize flaky e2e self-monitoring tests (#1125) (@k15r)
- 13204a3: fix: Unify CRs' descriptions in Kyma dashboard (#1095) (@mmitoraj)
Documentation
- e5de3ad: docs: Enhance Dynatrace guide for metric scenario (#1087) (@a-thaler)
- a9dfff4: docs: Fix example in the Dynatrace guide (#1137) (@skhalash)
- 8cc1113: docs: Improve condition messages and reasons (#1099) (@TeodorSAP)
- 03f16c0: docs: Leader Receiver Creator ADR (#1100) (@skhalash)
- bc28bb9: docs: Update pull-request-template.md (#1135) (@lilitgh)
Test suites
Contributors
k15r, chrkl, and 7 other contributors
Assets 5
1.15.1
1.16.1
1.16.0
Release Notes
General
- For the Telemetry Kubernetes API (CRD), we've aligned the status conditions, so that all reasons are named consistently (#1038).
- The Telemetry module is now visualized in the Kyma Dashboard (#1092).
Metrics
- Metrics provided using a MetricPipeline input now have a well-defined OTEL instrumentationScope attribute defined to identify the Telemetry module's input as origin of the metric, like io.kyma-project.telemetry/prometheus (#973)
Logs
- Update of Fluent Bit to version 3.0.3 (#1027)
Changelog
Features
- 7abb1c3: feat: Add instrumentation scope for metrics (#1041) (@rakesh-garimella)
- 2b0087c: feat: Add support to configure metric gateway replicas in Kyma Dashboard (#1042) (@chrkl)
- 514d676: feat: Enable sidecar injection for Fluent Bit by default (#1052) (@skhalash)
- 9e4f63d: feat: Instrument self-monitoring prober (#1072) (@skhalash)
- c10378e: feat: Synchronize CRs Reasons (#1038) (@TeodorSAP)
- 2cae1dc: feat: Update fluent-bit image version to 3.0.3 (#1027) (@hisarbalik)
Bug fixes
- d35811d: fix: Always set OwnerReference for telemetry-fluent-bit-sections ConfigMap (#1084) (@shorim)
- acc3198: fix: Update telemetry-manager to latest fluent-bit version 3.0.3 (#1082) (@a-thaler)
- 56ae4db: fix: Use different UI path for telemetry module view to not clash with Istio resources (#1092) (@a-thaler)
Documentation
- 50d6183: docs: Add self-monitor to architecture figures (#1045) (@chrkl)
- 854e8c3: docs: Fix error in self monitor + adjust diagrams (#1067) (@NHingerl)
- a709d75: docs: Fix wrong list type in documentation (#858) (@k15r)
- 31e0f79: docs: Improve license warning for loki integration guide (#1054) (@a-thaler)
- 2b56af7: docs: Update cloud logging container runtime dashboard (#1032) (@a-thaler)
- 5a85a4b: docs: Update release process documentation (#1053) (@TeodorSAP)
Test suites
- e2b14d7: test: Add instrumentation scope e2e (#1055) (@rakesh-garimella)
- 2dd309e: test: Add log pipeline multi pipeline e2e test (#1036) (@hisarbalik)
- 82d3264: test: Convert new OTel Collector basic auth/custom header/mtls E2E tests to unit tests (#1063) (@skhalash)
- e2d30ed: test: Do not execute self-monitor istio tests in Gardener suite (#1066) (@skhalash)
- 01bccd2: test: Fix flaky e2e test, test max pipelines in an isolated environment (#1061) (@hisarbalik)
- cd8e1da: test: Improve check for existence of OwnerReference in e2e tests (#1062) (@shorim)
- a54ee06: test: Migrate logs env tests to e2e tests (#1076) (@shorim)
- 94bee57: test: Migrate traces and metrics env tests to e2e tests (#1057) (@shorim)
- d107694: test: Negative self-monitoring LogPipeline E2E tests (#1033) (@skhalash)
- 20141e4: test: Remove env tests for LogParser (#1071) (@shorim)
Contributors
k15r, chrkl, and 7 other contributors
Assets 5
1.15.0
Release Notes
General
- All pipeline types verify if the configured TLS client certificate matches the configured TLS private key. If it does not match it reports an error in the pipeline status. (#1018)
Traces/Metrics
- Update of Otel Collector to version 0.99.0 (#1016)
Traces
- Removed OpenCensus integration and service endpoint which was used for the Istio trace integration. As Istio got migrated to OTLP, the endpoint has no usage any longer and got removed without replacement (#739)
Changelog
Features
- 3bbc4d9: feat: Adjusted the FlowHealthy status condition messages (#1014) (@skhalash)
- 8918dcb: feat: Clean up OpenCensus (#1022) (@TeodorSAP)
- 08e78ab: feat: Update open-telemetry image version to 0.99.0 (#1016) (@hisarbalik)
- b3c8bbc: feat: Update to prometheus 2.45.5 in self-monitor (#1035) (@a-thaler)
Bug fixes
- b556789: fix: Always clear conditions list for TracePipelines, LogPipelines, and LogParsers in first reconciliation (#1004) (@shorim)
- 2a2d9b5: fix: Check cert and key pair (#1018) (@rakesh-garimella)
Documentation
- c793a03: docs: Add ADR for module status metrics (#995) (@chrkl)
- 4a54f89: docs: Fix dynatrace integration example snippet (#1025) (@a-thaler)
- 5decf99: docs: Fix inconsistencies (#956) (@NHingerl)
Test suites
- deace2f: test: Fix flakiness by isolating test resources (#1019) (@skhalash)
- f606551: test: Improve unit test coverage for status (#1015) (@rakesh-garimella)
- accd307: test: Migrate Telemetry env tests to e2e tests (#1031) (@shorim)
- 1c5b9ac: test: Simplify log producer builder (#1026) (@skhalash)
- e0c459f: test: TLS cert validation e2e tests for logs/traces/metrics (#1007) (@rakesh-garimella)
Contributors
chrkl, skhalash, and 6 other contributors
Assets 5
1.14.0
Release Notes
General
- To use the memory limits more efficiently, the resource setup has been improved for all components. (#804)
- Support for Kubernetes 1.29 (#967)
Logs
- Improved status message for LogPipeline if the used certificate is invalid (#996)
Metrics/Traces
- You can now easily detect whether there is an invalid or expired client certificate, because the TracePipeline and MetricPipeline status shows what's wrong (#930).
Experimental
- Data flow problems in a LogPipeline are now reflected in the status for LogPipelines as well (#963)
Changelog
Features
- a30db4e: feat: Add GOMEMLIMIT configuration for all telemetry components (#964) (@hisarbalik)
- abdd26b: feat: Add TLS cert validation for metrics and traces (#999) (@rakesh-garimella)
- 2bdf3ac: feat: Introduce a pipeline_name label to simplify pipeline matching in self-monitor (#975) (@skhalash)
- 0d0a49a: feat: LogPipeline flow health status based on Fluent Bit alerts (#963) (@skhalash)
- 4c9cd72: feat: Reflect LogPipeline flow health status in the Telemetry status (#994) (@skhalash)
- 7729fb5: feat: Whitelist metrics scraped by self-monitor (#982) (@skhalash)
Bug fixes
- 69dadbc: fix: Flakiness in the logppipeline tests (#960) (@rakesh-garimella)
- aae3531: fix: Improve TLS cert validation for LogPipeline (#996) (@rakesh-garimella)
- ef4f9fc: fix: Invocation order for setting the status (#959) (@rakesh-garimella)
Documentation
Test suites
Contributors
chrkl, skhalash, and 3 other contributors
Assets 5
1.13.1
1.13.0
Release Notes
Logs
- You can now easily detect whether there is an invalid or expired client certificate, because the LogPipeline status shows what's wrong (#653).
Metrics/Traces
- Update to Otel Collector 0.97.0 (#923)
Experimental
The first version of the advanced pipeline status is available in the development release (#425), covering the following features:
- Collection of agent metrics with a new self-monitor component
- Calculation of alert situations in the self-monitor regarding data flow problems like backpressure
- Enrichment of pipeline status with alert situations
Changelog
Features
- 0c1b1be: feat: Add LogPipeline TLS Certificate Validation (#914) (@hisarbalik)
- bf93b9c: feat: Add self-monitor Fluent Bit alert rules (#948) (@skhalash)
- f24cea3: feat: Trigger immediate reconciliation when self-monitor alert state changes (#912) (@skhalash)
- 987b479: feat: Update to otel-collector 0.97. with updated resty lib (#923) (@a-thaler)
Bug fixes
- 999330a: fix: Certificater/key should be sanitized always (#953) (@rakesh-garimella)
- 39dd83a: fix: Set content-security-policy header in webhook handler (#949) (@chrkl)
Documentation
- 0e7c063: docs: Add ADR for LogPipeline health status (#934) (@chrkl)
- 757bfe4: docs: Fix unordered lists (#933) (@Disper)
Test suites
- c9b5db0: test: Add Gitleaks workflow (#941) (@chrkl)
- 02ee5cc: test: Allow self-monitor to be deployed in dev mode (#915) (@rakesh-garimella)
- b45a9d8: test: Migrate noisy span filters test to telemetrygen (#945) (@skhalash)
- c52d012: test: Migrate to telmetry gen for metrics (#940) (@rakesh-garimella)
- 3c64d30: test: Self-monitor webhook E2E test (#932) (@skhalash)
- e6dfed3: test: Telemetry OTLP-based components error/warning logs (#931) (@TeodorSAP)
Contributors
chrkl, Disper, and 5 other contributors
Assets 5
1.12.0
Release Notes
- Update Otel Collector to 0.96.0 (#882)
Changelog
Features
- 4eb82e2: feat: Add default status message to the self-monitor conditions (#902) (@rakesh-garimella)
- cbbf53d: feat: Add scrape config for self-monitor (#876) (@rakesh-garimella)
- a3792f8: feat: Ensure that auto-generated CRDs for development variant are always up-to-date (#899) (@shorim)
- 9b8899b: feat: Implement flow health probing (#868) (@skhalash)
- 4fbb215: feat: Introduce v1beta1 API for experimental channel (#869) (@shorim)
- 1f989f4: feat: Reflect flow health probes in Pipeline status conditions (#892) (@skhalash)
Bug fixes
- 6ef6e90: fix: Ensure pending condition is always preserved (#887) (@shorim)
- 9d563c5: fix: Explicitly set previous release tag to fix the auto-generated changelogs (#886) (@hisarbalik)
- ce390bf: fix: Remove unnecessary list (#861) (@k15r)
- d02ae7d: fix: Self-monitor alert matching (#901) (@skhalash)
- 7ff7c93: fix: Suppress otel-collector warning at startup by using localhost as default (#913) (@a-thaler)
Documentation
- b887d09: docs: Add documentation about access logs PoC (#894) (@chrkl)
- 4b85694: docs: Apply markdowlint rules in the Telemetry repository (#916) (@grego952)
- ab613ee: docs: Fix Dynatrace integration guide (#885) (@skhalash)
- 98a106d: docs: Remove unnecessary order for test-output documentation (#860) (@k15r)
Test suites
Contributors
k15r, chrkl, and 6 other contributors
Assets 5
1.11.1
Release Notes
A bug got fixed which was introduced with 1.11.0 and which was breaking the compatibility of the TracePipeline status attribute (#891)
Changelog
Bug fixes
- 4adeac3: fix: Ensure pending condition is always preserved (#891) (@hisarbalik)
Contributors
hisarbalik