You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Describe the bug
If trying to run more than one rule pod there doesn't appear to be a way to specify an argument with '--label=replica="podhostname"'. This is needed to run Thanos rule as in HA mode.
Steps to reproduce the issue:
Set replicaCount > 1 in the rule section of the Thanos chart
Expected behavior
Chart should add an argument to each pod.
--label=replica="-rule-0"
Screenshots
NA
Additional context
Thanos Chart version 0.3.27
arguments from kubectl describe of one of my rule pods
The text was updated successfully, but these errors were encountered:
grabango-joonho
changed the title
Thanos Rule: need ability to add label with value as hostname
Thanos Rule: when running multiple rule pods can't specify an argument with hostname of pod as label value
Oct 2, 2020
Describe the bug
If trying to run more than one rule pod there doesn't appear to be a way to specify an argument with '--label=replica="podhostname"'. This is needed to run Thanos rule as in HA mode.
Steps to reproduce the issue:
Set replicaCount > 1 in the rule section of the Thanos chart
Expected behavior
Chart should add an argument to each pod.
--label=replica="-rule-0"
Screenshots
NA
Additional context
Thanos Chart version 0.3.27
arguments from kubectl describe of one of my rule pods
Containers:
thanos-rule:
Container ID: docker://3a7b7d9b368d3d11eb5df21761b04847d3ab3044a0ff96efea4add166af8af7b
Image: quay.io/thanos/thanos:v0.14.0
Image ID: docker-pullable://quay.io/thanos/thanos@sha256:06f4cd5175467865249a91fa20b7985976bd98e1b10502e360ba565851b5b286
Ports: 10902/TCP, 10901/TCP
Host Ports: 0/TCP, 0/TCP
Args:
rule
--data-dir=/var/thanos/store
--log.level=debug
--log.format=logfmt
--http-address=0.0.0.0:10902
--grpc-address=0.0.0.0:10901
--objstore.config-file=/etc/config/object-store.yaml
--rule-file=/etc/rules/*.yaml
--resend-delay=60m
--alertmanagers.url=http://alertmanager-operated.monitoring.svc.cluster.local:9093
--alert.query-url=https://alertmanager.mydomain.com
--query=my-release-query-http.my-release-namespace.svc.cluster.local:10902
--label=cluster="thanos-cluster"
The text was updated successfully, but these errors were encountered: