Version | Image address | Description | Release date | Impact |
V1.1.11 | | The service degradation feature is provided. If a remote write fails, data collection on the primary link stably continues. The Global Config in collection jobs can be modified. The Remote Write feature is enhanced. If you fail to resolve a domain name by using CoreDNS, the cached IP address corresponding to the domain name is sent. Remote Write allows you to configure multiple IP addresses.
| December 2022 | No impact on workloads. |
V1.1.9 | | The Prometheus agent supports the following CPU architectures: AMD64, ARM, ARM64, ppc64le, and s390x. The self-monitoring capabilities of the Prometheus agent are enhanced. The garbage collection (GC) policy of the Prometheus agent is optimized. The multi-replica Target scheduling policy is optimized to prevent memory leaks of Worker nodes. The memory degradation issue of the Prometheus agent is fixed. The deadlock of boundary conditions in multi-replica scenarios is fixed. The following service discovery capabilities are provided: IONOS, PuppetDB, Uyuni, and Vultr.
| September 2022 | No impact on workloads. |
V1.1.7 | | The metrics and dashboards for the latest GPU Exporter are supported. For more information, see Enable GPU monitoring for an ACK cluster. | July 2022 | No impact on workloads. |
V1.1.6 | | The data collection issue of GPU Exporter V1.x is fixed. | June 2022 | No impact on workloads. |
V1.1.5 | | The integration center is supported. Clusters that contain more than 10,000 nodes are supported. ServiceMonitor and PodMonitor configurations that are not created in the Managed Service for Prometheus console can be synchronized to Managed Service for Prometheus. ServiceMonitor and PodMonitor configurations that declaratively specify how groups of services are monitored and are not created in the Managed Service for Prometheus console can be discovered by using the service discovery feature. The maximum number of agent HPA replicas can be configured by using parameters. The fields related to the Job metric of Managed Service for Prometheus can be edited. The configuration files related to ServiceMonitor, PodMonitor, and Prometheus.yaml can be verified. CPU, memory usage, and system stability are optimized.
| May 2022 | No impact on workloads. |
V1.1.4 | | | April 2022 | No impact on workloads. |
V1.1.3 | | Helm is compatible with the Kubernetes version 1.22 of ACK clusters. | February 2022 | No impact on workloads. |
V1.1.2 | | kube-state-metrics V2.3.0-755434c-aliyun is updated. | January 2022 | No impact on workloads. |