71 lines
2.8 KiB
Markdown
71 lines
2.8 KiB
Markdown
---
|
|
title: "Operations"
|
|
linkTitle: "Operations"
|
|
weight: 40
|
|
description: >
|
|
Operational guides for deploying, monitoring, and maintaining the Edge Developer Platform components.
|
|
---
|
|
|
|
|
|
## Operations Overview
|
|
|
|
This section covers operational aspects of the Edge Developer Platform. In general there is no operation - it's just monitoring and fixing in an developer mode of operation.
|
|
|
|
## Deployments
|
|
|
|
### EDP
|
|
|
|
EDP is running on two OTC clusters (remember: this just means that we twice run the [infra-deploy pipeline](https://edp.buildth.ing/DevFW/infra-deploy/actions?workflow=deploy.yaml&actor=0&status=0) as eyerything is code!)
|
|
|
|
|
|
|
|
#### Further references for infrastructural informations
|
|
|
|
* OTC:
|
|
* [IPCEI-CIS Confluence](https://confluence.telekom-mms.com/spaces/IPCEICIS/pages/1000105031/OTC)
|
|
|
|
### Edge Connect
|
|
|
|
The Edge and Orka clouds in Edge Connect are in the perspective of eDF just deployment targets of EDP applications. Thus we are user of them. Physically both are Gardener Kubernetes clusters.
|
|
|
|
|
|
Basically it is intended for users just to use the web ui: https://hub.apps.edge.platform.mg3.mdb.osc.live
|
|
|
|
|
|
|
|
#### Further references for infrastructural informations
|
|
|
|
|
|
|
|
But we also got access on the cluster level for operations issues, see picture above. How to get access is described in the following links:
|
|
* [IPCEI-CIS Confluence](https://confluence.telekom-mms.com/spaces/IPCEICIS/pages/1122869593/Edge+Cloud)
|
|
* [IPCEI-CIS Jira](https://jira.telekom-mms.com/browse/IPCEICIS-6222?focusedId=3411527&page=com.atlassian.jira.plugin.system.issuetabpanels%3Acomment-tabpanel#comment-3411527)
|
|
* Hint: Get the kubeconfig of the `platform` in your Gardener Account Settings, then run ```gardenctl target --garden mg3 --project platform --shoot edge``` to authenticate and get the cluster kubectl context.
|
|
|
|
## Monitoring & Observability
|
|
|
|
On EDP the observability cluster is meant to monitor the platform stacks, e.g. by [Grafana](https://grafana.observability.buildth.ing).
|
|
But there is no operational monitoring lifecycle in place. we didn't define metrics or alerts as there is no operational mode yet.
|
|
|
|
Most monitoring and observability is done through grafana, deployed through the observability stack.
|
|
|
|
Login is found in the `grafana-admin-credentials` secret.
|
|
|
|
|
|
NOTE document that default deployed stacks are different depending on is_observability flag
|
|
|
|
|
|
|
|
## Maintenance
|
|
|
|
We maintain EDP on an per issue driven strategy.
|
|
|
|
### Updates & Upgrades
|
|
|
|
We update occassionally per [component](../components/), which either can be a [platform application](../components/orchestration/stacks/) or an [orchestration pipeline](../components/forgejo/actions/actions.md).
|
|
|
|
### Backup & Recovery
|
|
|
|
EDP Customer data is backuped, see https://jira.telekom-mms.com/browse/IPCEICIS-5017
|
|
|
|
|