From ea3923abe224c11e5732e5d982778920e6eb407b Mon Sep 17 00:00:00 2001 From: Dan Fuchs Date: Thu, 9 Jan 2025 18:09:47 -0600 Subject: [PATCH] DM-48382: sasquatch app metrics - use nightly telegraf [telegraf](https://github.com/influxdata/telegraf) has still not released the [feature to allow union fields to be specified as tags](https://github.com/influxdata/telegraf/commit/bcea9a28c0070b4e959afe25ccdcde2c44c937ae), which will let us get a `service` tag on some Gafaelfawr metrics. This is planned to be released in [v1.34.0](https://github.com/influxdata/telegraf/milestone/118). Until then, we'll use a nightly telegraf build from 2025-01-09 and pushed to our `lsst-sqre` github docker registry (ghcr.io): ``` docker pull quay.io/influxdb/telegraf-nightly:alpine docker tag quay.io/influxdb/telegraf-nightly:alpine ghcr.io/lsst-sqre/telegraf:nightly-alpine-2025-01-09 docker push ghcr.io/lsst-sqre/telegraf:nightly-alpine-2025-01-09 ``` --- applications/sasquatch/README.md | 4 ++-- applications/sasquatch/charts/app-metrics/README.md | 4 ++-- applications/sasquatch/charts/app-metrics/values.yaml | 9 +++++++-- 3 files changed, 11 insertions(+), 6 deletions(-) diff --git a/applications/sasquatch/README.md b/applications/sasquatch/README.md index 47dabedf26..b73b49cee9 100644 --- a/applications/sasquatch/README.md +++ b/applications/sasquatch/README.md @@ -97,8 +97,8 @@ Rubin Observatory's telemetry service | app-metrics.globalAppConfig | object | See `values.yaml` | app-metrics configuration in any environment in which the subchart is enabled. This should stay globally specified here, and it shouldn't be overridden. See [here](https://sasquatch.lsst.io/user-guide/app-metrics.html#configuration) for the structure of this value. | | app-metrics.globalInfluxTags | list | `["application"]` | Keys in an every event sent by any app that should be recorded in InfluxDB as "tags" (vs. "fields"). These will be concatenated with the `influxTags` from `globalAppConfig` | | app-metrics.image.pullPolicy | string | `"Always"` | Image pull policy | -| app-metrics.image.repo | string | `"docker.io/library/telegraf"` | Telegraf image repository | -| app-metrics.image.tag | string | `"1.30.2-alpine"` | Telegraf image tag | +| app-metrics.image.repo | string | `"ghcr.io/lsst-sqre/telegraf"` | Telegraf image repository | +| app-metrics.image.tag | string | `"nightly-alpine-2025-01-09"` | Telegraf image tag | | app-metrics.imagePullSecrets | list | `[]` | Secret names to use for Docker pulls | | app-metrics.influxdb.url | string | `"http://sasquatch-influxdb.sasquatch:8086"` | URL of the InfluxDB v1 instance to write to | | app-metrics.nodeSelector | object | `{}` | Node labels for pod assignment | diff --git a/applications/sasquatch/charts/app-metrics/README.md b/applications/sasquatch/charts/app-metrics/README.md index 3f75c1aa74..b00e09fd45 100644 --- a/applications/sasquatch/charts/app-metrics/README.md +++ b/applications/sasquatch/charts/app-metrics/README.md @@ -16,8 +16,8 @@ Kafka topics, users, and a telegraf connector for metrics events. | globalAppConfig | object | See `values.yaml` | app-metrics configuration in any environment in which the subchart is enabled. This should stay globally specified here, and it shouldn't be overridden. See [here](https://sasquatch.lsst.io/user-guide/app-metrics.html#configuration) for the structure of this value. | | globalInfluxTags | list | `["application"]` | Keys in an every event sent by any app that should be recorded in InfluxDB as "tags" (vs. "fields"). These will be concatenated with the `influxTags` from `globalAppConfig` | | image.pullPolicy | string | `"Always"` | Image pull policy | -| image.repo | string | `"docker.io/library/telegraf"` | Telegraf image repository | -| image.tag | string | `"1.30.2-alpine"` | Telegraf image tag | +| image.repo | string | `"ghcr.io/lsst-sqre/telegraf"` | Telegraf image repository | +| image.tag | string | `"nightly-alpine-2025-01-09"` | Telegraf image tag | | imagePullSecrets | list | `[]` | Secret names to use for Docker pulls | | influxdb.url | string | `"http://sasquatch-influxdb.sasquatch:8086"` | URL of the InfluxDB v1 instance to write to | | nodeSelector | object | `{}` | Node labels for pod assignment | diff --git a/applications/sasquatch/charts/app-metrics/values.yaml b/applications/sasquatch/charts/app-metrics/values.yaml index 97b4e6c643..a25f46bf14 100644 --- a/applications/sasquatch/charts/app-metrics/values.yaml +++ b/applications/sasquatch/charts/app-metrics/values.yaml @@ -41,12 +41,17 @@ cluster: name: sasquatch # These values refer to the Telegraf deployment and config +# We are currently using a nightly build to get functionality that will be +# released in 1.34, in March 2025. Once 1.34 has been released, we should +# stop using this nightly build. image: # -- Telegraf image repository - repo: "docker.io/library/telegraf" + repo: "ghcr.io/lsst-sqre/telegraf" + #repo: "docker.io/library/telegraf" # -- Telegraf image tag - tag: "1.30.2-alpine" + tag: "nightly-alpine-2025-01-09" + # tag: "1.30.2-alpine" # -- Image pull policy pullPolicy: "Always"