Important news and breaking changes
-
NOTE: As of Prometheus 3.0, UTF-8 strings are valid for metric names and label names. However, for backward compatibility, this release of YACE still uses the old, stricter legacy validation scheme. UTF-8 validation will be enabled in a feature version of YACE, thus requiring that your remote destination is compatible with UTF-8 support.
-
BREAKING CHANGE: the AWS SDK v2 is now the default in YACE. Use the flag
aws-sdk-v1
to switch back to SDK v2. Flagaws-sdk-v2
has been removed. -
NEW FEATURE:
exportAllDataPoints
, enables the inclusion of past metric data points from the CloudWatch response if available.
- [CHANGE] Make aws sdk v2 the default choice by @cristiangreco
- [FEATURE] Support history data export by @woehrl01
- [FEATURE] Add AWS/Transfer as available service by @thepalbi
- [FEATURE] Add auto-discovery for Directory Services(MicrosoftAD) by @RuslanMustaev
- [FEATURE] Add support for Redshift-Serverless by @nickbazinet
- [FEATURE] Add db connections avg panel to RDS dashboard by @yduartep
- [FEATURE] Add example for lambda_edge by @tyagian
- [FEATURE] sagemaker: additional InferenceComponent support by @tristanburgess
- [ENHANCEMENT] Update Go version by @SuperQ
- [ENHANCEMENT] Use Prometheus common version library by @SuperQ
- [ENHANCEMENT] Update container repositories by @SuperQ
- [ENHANCEMENT] Speed up build metric name by @jeschkies
- [ENHANCEMENT] Add guard to hot logging location in associator by @thepalbi
- [ENHANCEMENT] Update resource association logic to try both with and without dimension fixes by @tristanburgess
- [ENHANCEMENT] Change discovery runtime model field from Type -> Namespace by @kgeckhart
- [BUGFIX] Fix
CachingFactory
concurrent usage issues by @andriikushch - [BUGFIX] Correctly run tests in CI and fix failing tests by @jeschkies
- [BUGFIX] Fix doc about non-existing
debug
flag by @zipkid - [BUGFIX] Update URL to Helm Chart in docs by @koralowiec
- [BUGFIX] Add missing license header to
associator_logging_test.go
by @cristiangreco - [BUGFIX] Dashboards: replace
scrape_job
label withjob
by @yduartep - [BUGFIX] RDS dashboard: use average for cpu utilization to align with AWS best practices by @yduartep