Bump kube-prometheus-stack for multi-cluster support #162
Conversation
VineethReddy02
force-pushed
the
vineeth-multi-cluster-support
branch
from
3c53d03 to
12d5a1e
Compare
cli/cmd/upgrade/upgrade.go
Outdated
| "https://raw.githubusercontent.com/prometheus-operator/prometheus-operator/v0.47.0/example/prometheus-operator-crd/monitoring.coreos.com_servicemonitors.yaml", | ||
| "https://raw.githubusercontent.com/prometheus-operator/prometheus-operator/v0.47.0/example/prometheus-operator-crd/monitoring.coreos.com_thanosrulers.yaml", | ||
| "https://raw.githubusercontent.com/prometheus-operator/prometheus-operator/v0.47.0/example/prometheus-operator-crd/monitoring.coreos.com_prometheusrules.yaml", | ||
| "https://raw.githubusercontent.com/prometheus-operator/prometheus-operator/v0.49.0/example/prometheus-operator-crd/monitoring.coreos.com_alertmanagerconfigs.yaml", |
There was a problem hiding this comment.
is there anyway we can have a test to make sure this version is aligned with the kube-prometheus-stack version? Otherwise this is very easy to forget to do.
There was a problem hiding this comment.
Tests definitely sound reasonable. But usually, during the kube-prometheus version bump, we read the upgrade guide in kube-prometheus helm chart to understand the changes for upgrades.
I will add the tests on this.
| @@ -0,0 +1,128 @@ | |||
| # Tobs helm values config | |||
There was a problem hiding this comment.
hmm. I think there are two different types of config here.
- config which you need only for things that don't use the CLI (high-availability)
- config applicable to both helm chart and CLI. (multicluster)
Type 1 is ok here. I'd move type 2 stuff into a new top-level /docs directory. It does not belong under /chart and I'd add new links from the front-page readme here.
There was a problem hiding this comment.
So currently using CLI we offer only abstract features with flags like HA, integration with forge, etc... all these features can also be configured within values.yaml. All the flexibility lies in the values.yaml so the intention behind creating this file is to document all possible configurations across different components in tobs.
Yes, I agree with you only config related to values.yaml will reside in this file. Anything about CLI will be in a new file in the high level root directory of this repo.
chart/docs/values-config.md
Outdated
| externalLabels: | ||
| cluster: <clusterName> | ||
| remoteRead: | ||
| - url: "<PROMSCALE_ENDPOINT>/read" |
There was a problem hiding this comment.
| - url: "<PROMSCALE_ENDPOINT>/read" | |
| - url: "<PROMSCALE_SERVICE_ENDPOINT_OF_DATA_AGGREGATION_CLUSTER>/read" |
| replicas: 3 | ||
| ``` | ||
|
|
||
| ## Multi-cluster support |
There was a problem hiding this comment.
can we have a diagram for this section showing the various clusters etc.
VineethReddy02
force-pushed
the
vineeth-multi-cluster-support
branch
from
12d5a1e to
f24b62f
Compare
VineethReddy02
force-pushed
the
vineeth-multi-cluster-support
branch
from
f24b62f to
8b7909a
Compare
|
|
||
| ## Multi-cluster support | ||
|
|
||
| <img src="./../../docs/assets/multi-cluster.png" alt="multi-cluster setup diagram" width="800"/> |
There was a problem hiding this comment.
I believe you forgot to add this image in the PR?
There was a problem hiding this comment.
Added it to the PR here is the file: https://github.com/timescale/tobs/pull/162/files#diff-1a7cd77999e454c5a336fcb163eb5108d70057981aa40d648c619671eba20b3e
2 participants
fixes: #156