# Agent

## 1.9.2 - 2022-02-02

* Adds pushgateway error handling when it is not available.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.9.2`

## 1.9.1 - 2021-11-25

* Fix missing DCGM-Exporter metrics.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.9.1`

## 1.9.0 - 2021-11-11

* Adds env `PROMETHEUS_GPU_METRICS_SOURCE` to specify the prometheus GPU metrics source. Options `cadvisor` for cAdvisor accelerator metrics or `dcgm` to use nvidia DCGM-exporter metrics. Default cadvisor
* Adds env `PROMETHEUS_HONOR_LABELS` set the prometheus honor\_labels scrape setting. Default false.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.9.0`

## 1.8.0 - 2021-07-22

* Handles all offline provider sync errors and retries syncing and pushing metrics once available
* Improves thanos UX for better error handling
* Removes env `PROMETHEUS_TSDB_BLOCK_DURATION` and defers to the thanos receive component alone for querier and store support

Docker Image: `docker.replex.io/replex/k8sagent-go:1.8.0`

## 1.7.20 - 2021-07-13

* Adds env `PROMETHEUS_TSDB_BLOCK_DURATION` for Thanos querier support

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.20`

## 1.7.19 - 2021-06-16

* Bug fixes and improvements

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.19`

## 1.7.18 - 2021-06-01

* Bug fixes and improvements

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.18`

## 1.7.17 - 2021-05-31

* Adds a new optional environment variable: `PROMETHEUS_BEARER_TOKEN` to set bearer token for prometheus server requests.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.17`

## 1.7.16 - 2021-05-18

* Fix bug in Alicloud detection

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.16`

## 1.7.15 - 2021-05-18

* Bug fixes and improvements

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.15`

## 1.7.14 - 2021-05-10

* Add Thanos support.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.14`

## 1.7.13 - 2021-02-15

* Add Alibaba cloud support.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.13`

## 1.7.12 - 2021-01-14

* Adds node CPU usage support for Stackdriver metric providers.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.12`

## 1.7.11 - 2020-12-07

* Sets the default Prometheus node label to `node` (previously `instance`).

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.11`

## 1.7.10 - 2020-11-10

* Bug fix in node CPU usage calculation.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.10`

## 1.7.9 - 2020-11-09

* We now use actual node CPU usage instead of the sum of the container CPU usage. This helps to account for CPU usage by non-container workload.
* Adds support for failed metrics synchronization on disk.
  * By setting the environment variable `METRICS_CACHE_DISK` to either `true/false`, you can specify whether the metrics are cached on disk or in memory. By default, `true`.
  * If metrics are cached on disk, you can specify the path on disk by setting the environment variable `METRICS_CACHE_DISK_DIR`. By default, `/data/metrics`.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.9`

## 1.7.8 - 2020-10-19

* Adds a new optional environment variable: `METRICS_RETRY_INTERVAL_SECONDS` to set interval between retry to push metrics.
* Adds metric caching if push to the pushgateway fails and auto-retries the push once per value from `METRICS_RETRY_INTERVAL_SECONDS` or 300 seconds by default.
* Adds two new metrics `replex_agent_retry_cache_size` and `replex_agent_failed_metrics_total`. First one stands for the amount of metrics that are waiting to be re-sent, second one represents the total count of once failed metric push.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.8`

## 1.7.7 - 2020-10-14

* Adds a new optional environment variable: `METRICS_FILESYSTEM` to specify the filesystem metric source. Options: `cadvisor`, `csi`.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.7`

## 1.7.6 - 2020-10-06

* Adds a new optional environment variable: `USE_CONTROL_PLANE_COST` to track costs of the Kubernetes Control Plane. Set to `true/false` to enable/disable the feature.
* Adds node local disk support for Datadog metric provider.
* Adds Daemon- and StatefulSet support for Instana.
* Adds the collection of node provider id metric.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.6`

## 1.7.5 - 2020-09-25

* Adds a new optional environment variable: `CLOUD_PROVIDER_OVERRIDE` to manually overwrite the cloud provider. Set to `aws, azure, gce, custom` choose the provider.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.5`

## 1.7.4 - 2020-09-18

* Disabled logs for pods created within the last 5 minutes and for already succeeded pods.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.4`

## 1.7.3 - 2020-09-16

* Adds a new optional environment variable: `PROMETHEUS_NODE_LABEL` to specify the label that represents the node in the Prometheus metrics.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.3`

## 1.7.2 - 2020-09-16

* Skip pod metrics if node is not ready yet.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.2`

## 1.7.1 - 2020-09-15

* Adds a new optional environment variable: `ONLY_USE_READY_NODES` to track only nodes that are in "Ready" state. Set to `true/false` to enable/disable the feature.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.1`

## 1.7.0 - 2020-09-07

* Adds node local disk support for Prometheus metric provider.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.7.0`

## 1.6.1 - 2020-08-19

* Renaming the label that represents the node in the Prometheus metrics from `node` to `instance`.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.6.1`

## 1.6.0 - 2020-07-28

* Adds Instana topology provider support.
* Adds a new optional environment variable: `KUBERNETES_INFO_PROVIDER` to set the topology provider. Set to `kubernetes, instana` to choose between Kubernetes and Instana.
* Adds a new optional environment variable: `INSTANA_CLUSTER_ID` to set the Instana cluster id.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.6.0`

## 1.5.0 - 2020-06-26

* Adds the collection of cluster and Prometheus version.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.5.0`

## 1.4.2 - 2020-06-02

* Datadog: using `kubernetes.memory.working_set` for container memory metrics
* Instana: using `memory.usage` of docker plugin for container memory metrics.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.4.2`

## 1.4.1 - 2020-05-27

* Renaming environment variable from `INSTANA_URL` to `INSTANA_BASE_URL`

Docker Image: `docker.replex.io/replex/k8sagent-go:1.4.1`

## 1.4.0 - 2020-05-25

* Adds support for Instana metrics provider.
* Skipping unmounted PVCs.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.4.0`

## 1.3.0 - 2020-02-29

* Adds support for Stackdriver \\(Google Operations\\) metrics provider.

Docker Image: `docker.replex.io/replex/k8sagent-go:1.3.0`