Skip to content

Commit a66a85f

Browse files
author
Martin Jackson
committed
Skeleton of detail page
1 parent a85393e commit a66a85f

File tree

1 file changed

+69
-3
lines changed

1 file changed

+69
-3
lines changed

content/patterns/ramendr-starter-kit/installation-details.adoc

Lines changed: 69 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -11,8 +11,74 @@ include::modules/comm-attributes.adoc[]
1111

1212
== Installation Steps
1313

14-
=== Major Validation Scripts
14+
.Installation Steps
15+
16+
The pattern will execute the following steps on the cluster:
17+
18+
. Apply Subscriptions and Applications to Hub Cluster
19+
. This includes ACM, ODF and ODF MultiCluster Operator on the hub cluster
20+
. Build managed clusters (ocp-primary and ocp-secondary) with Hive
21+
.. The managed clusters have identical configuration regarding Subscriptions and Applications, so they are both in the resilient clusterGroup
22+
. opp-policy app is responsible for copying CA certificates to the following places:
23+
.. Creating a configmap cluster-proxy-ca-bundle in namespace openshift-config
24+
.. Assigning this configmap to the proxy cluster resource
25+
.. Adding the certificate material to ramen-dr-cluster-operator config in openshift-dr-system
26+
. regional-dr app is responsible for:
27+
.. ensuring ODF is setup properly
28+
.. Installing submariner add-ons on managed clusters
29+
.. Creating DRPolicy, MirrorPeer, DRPC, and Placement objects for RamenDR
30+
.. Installing the VM workload on the primary cluster
31+
.. Disabling Sync on the regional-dr app to prevent potential conflicts later
32+
33+
=== Various Scripts included in the pattern and how to use them
34+
35+
* scripts/cleanup-gitops-vms-non-primary.sh
36+
37+
Designed to be run when you need to manually cleanup resources from a "failed" cluster.
38+
Intended to be run with the kubeconfig from the hub cluster; it will determine where to
39+
delete resources based on the current DRPC state.
40+
41+
* scripts/download-kubeconfigs.sh
42+
43+
Will download and extract the kubeconfigs for the managed clusters to the current directory.
44+
Useful when you need to check something or do something on one of the managed clusters.
45+
46+
* charts/hub/opp/scripts/argocd-health-monitor.sh
47+
48+
Ensures that ArgoCD is progressing properly in deploying resources. A workaround for an ArgoCD
49+
bug we ran into during development.
50+
51+
* charts/hub/opp/scripts/odf-ssl-precheck.sh
52+
53+
Ensures all the preconditions have been met for extracting certificates to distribute among the
54+
clusters.
55+
56+
* charts/hub/opp/scripts/odf-ssl-certificate-extraction.sh
57+
58+
This script does the actual work of extracting and distributing the CA material to the various
59+
places it needs to go. Will also restart velero (OADP) pods if needed.
60+
61+
* charts/hub/rdr/scripts/odf-dr-prerequisites-check.sh
62+
63+
Ensures that ODF is fully ready to be configured for Disaster Recovery. In particular waits for ODF
64+
to finish deployment and for the NooBaa/S3 service to be operational on all clusters.
65+
66+
* charts/hub/rdr/scripts/submariner-prerequisites-check.sh
67+
68+
Ensures that submariner is running properly and operational on both clusters. This is required for
69+
ODF PVC replication to work.
70+
71+
* charts/hub/rdr/scripts/edge-gitops-vms-deploy.sh
72+
73+
This script deploys the VM workload to the primary cluster. It uses the Validated Patterns helm chart but
74+
is not an argo application to avoid starting up resources on clusters where we do not want them running.
75+
Thus it runs from the Hub cluster.
76+
77+
* charts/hub/rdr/scripts/drpc-health-check-argocd-sync-disable.sh
78+
79+
This script disables sync on the rdr application to prevent ArgoCD from changing something during the
80+
failover process.
81+
82+
1583

16-
== Triggering a Failover
1784

18-
=== How to Run Manual Cleanup

0 commit comments

Comments
 (0)