opni
opni copied to clipboard
rancher
Multi Cluster Observability with AIOps
Add a Sync Task to check when the logging cluster is installed that a valid OpenSearch Destination object exists that points to the Opni alerting Destination proxy
HTTP proxy for open search opni webhook destination. Requires TLS, so client certs will have to be mounted in OpenSearch.
Timeline API return types https://github.com/rancher/opni/blob/a542630c95138e7005747ca4eb4c558e16608921/pkg/apis/alerting/v1/alerting.condition.proto#L480-L493 now have an additional field that contains an array of fingerprints: Sample output { "items": { "Ak8vDnrUekfCgWvzKb4SBP": { "windows": [ { "start": "2023-03-08T19:08:18.672644298Z", "end": "2023-03-08T19:10:08.662919079Z",...
Hovering on alarm status no longer shows the status reason
Need to have a snapshot/restore process for the gateway KV store. Questions: - Should this be storage specific or agnostic of the gateway store - Add to the storage interface?
With the release of nats-py v2.3.0 which supports object storage, update the Opni nats wrapper to incorporate this change.
Currently, within the Opensearch Opni plugin, there are only 3 sections for "Insights", "Events" and "Log Templates". The "Clusters" panel is missing.
To allow statefulset rollouts to work correctly we should have readiness checks.
It appears that the drain service is attempting to create a bucket that already exists. Here are the relevant logs: ``` 2023-05-26 17:23:49,669 - INFO - Connected to S3 client...
Template is currently : ```js helm -n ${ this.namespace } install opni-agent-crd ${ imageCrd } --create-namespace && helm -n ${ this.namespace } install opni-agent ${ imageMain } ${ prometheus }...