Wednesday, September 4, 2024

Oracle KeyVault 21.9 Multimaster setup steps

Oracle KeyVault Multimaster setup steps


First node IP : 192.168.194.122 | voracle8x-oracle-key-vault2

Second node IP : 192.168.194.123 | voracle8x-oracle-key-vault3


Note: the DNS and NTP are manadtory for the multimaster setup. Since I didnt have both available; I didnt try making /etc/hosts in first attempt to honour oracle's requirement not to tamper any setting in the server and use only web UI or restapi or okvutil to do tasks with the machine.


Prereq: Atleast make a NTP server entry in system tab (even if the NTP server isnt reachable from the OKV node 1)


Controller node creation:


1. Backup the existing nodes 

Use backup & restore from the web UI.

workaround: I used the virtual box vm snapshot instead in vm shut state.


2. Once remote backup finishes; go to cluster tab


3. Verify server IP


4. Select "YES" for First node of cluster option


5. Enter node name, this cant be changed later


6. Enter cluster name


7. Enter cluster subgroup name


8. Convert the node to candidate node (eventually this is the first node or controller node in the cluster).


In our case:


Clustername: voracle-okv-cl01


cluster subgroup name: voracle-okv-cl01-sg01


Issues: The OKV cluster doesnt allow the conversion without NTP server; to mitigate it we tried fixing the port forward for internet. No help, the NTP server is still not reachable. 

Finally we entered the public NTP server 0.uk.pool.ntp.org; the OkV server failed to reach the ntp server. 

But still the conversion worked.



Induction completed until here:


First node IP : 192.168.194.122 | voracle8x-oracle-key-vault2



Pending:


Second node IP : 192.168.194.123 | voracle8x-oracle-key-vault3


Candidate node addition (2nd node in read write pair):


Prereqs: Backup the existing controller or first node.


Candidate node addition steps:

Backup steps:

1. shutdown the controller node

2. Take a snapshot of the controller node from virtual box

3. So the backups are

A. Backup as of node1_snap_04Sep2024_controller_success & node2_snap_04sep2024_postconfig_premultimaster


Oracle VM VirtualBox Manager


Oracle VM VirtualBox Manager


1. Login to web UI console as system admin user

http://192.168.194.122 [controller node]

2. Goto system tab

3. click on backup & restore ---- -> backup is already secured

4. Ensure a remote backup is secured---- -> backup is already secured

5. Firewall ports are opened as mentioned in OKV guide (in local lan, it shouldnt be concern) ---> same VLAN and no firewall between

6. Goto cluster tab

7. Click on management tab from left navigation bar

8. Click add under cluster details

9. Add cluster node details to the "add candidate node section"

10. enter the recovery passphrase of controller node

11. Add candidate node section details:

a. Add as read/write pair : yes

b. Node id: unique id auto populated (leave it as is)

c. Node name: enter unique name for the candidate node; again auto populated with server name

d. cluster subgroup: keep the original here, since the DC is same and the servers are build only for HA not DR

e. cluster name: auto populated

f. IP address

g. Certificate of the candidate node:

i. open a new browser window with system admin credential; login to

https://192.168.194.123 [candidate node]

ii. First node of the cluster: No

iii. Enter recovery passphrase from controller node

iv. IP of the controller node

v. certificate of the controller node: on bottom of the original screen open already - you should see the certificate available, just copy the certificate content

vi. come back to the candidate browser node: paste the certificate from controller node

vii. now after validation of recovery passphrase and certificate, click convert to candidate node

viii. once conversion is complete, a new certificate appears on this page



h. copy this certificate to controller node browser now, paste under candidate node certificate section

i. Click add node




Progress track:

>>> it will take sometime (upto an hour or more based on volume of data to sync) to finish

>>>once it finishes, all the nodes will be displayed as active nodes

Cluster Management - Google Chrome


From other node:

Cluster Management - Google Chrome

Cluster Monitoring - Google Chrome


>> secure an offline vm snap at this stage.

node1: node1_snap_04Sep2024_post_multimaster_setup

node2: node2_snap_04Sep2024_post_multimaster_setup


YouTube video:




Thanks


No comments:

Post a Comment

Flashback data archive steps

 Objective: Ways to track DML changes in a table Detailed objective: We should be able to track the changes in the table at any point in tim...