ExaCC: Scale VM Cluster Exadata Usable Storage

 

- Overview:

• With multiple VMs per Exadata Infrastructure feature release, you can scale up or down the following resources on VM cluster:
• Local storage /u02. Scaling memory requires a rolling restart of the cluster nodes to take effect.
• Memory. Scaling memory requires a rolling restart of the cluster nodes to take effect.
• OCPU. Scaling OPCU is Online.
• Exadata storage. Scaling OPCU is Online.
• The minimum recommended size is 2 TB.
• Use the following formula to calculate the minimum required ASM storage:
• For each disk group, for example, DATA, RECO, note the total size and free size by running the "asmcmd lsdg" command on any node of the VM cluster.
• Calculate the used size as (Total size - Free size) / 3 for each disk group. The /3 is used because the disk groups are triple mirrored.
• DATA:RECO ratio is:
• 80:20 if Local Backups option was NOT selected in the user interface.
• 40:60 if Local Backups option was selected in the user interface.
• Ensure that the new total size as given in the user interface passes the following conditions:
• Used size for DATA * 1.15 <= (New Total size * DATA % ).
• Used size for RECO * 1.15 <= (New Total size * RECO % ).

In this blog, I'll demonstrate the steps to scale up the usable Exadata storage using OCI console.

Steps to Scale Up VM Cluster Exadata Usable Storage 

1. Note the current total size and free size by running "asmcmd lsdg" command on any node of the VM cluster.

2.  Sign in to your OCI tenancy where your Exadata Database Service on Cloud @ Customer system is deployed.

3. Navigate to "Oracle Database" > "Oracle Exadata Database Service on Cloud@Customer" > "Exadata VM Clusters".

4. Click on your VM cluster name.

5. On VM cluster details home page, click "Scale VM Resources" button.

  

6. On "Scale VM Cluster" screen, under "Configure the Exadata storage" section, specify the usable Exadata storage new size then click "Save Changes" button. The new size will be distributed based on DATA:RECO ratio, which is 80:20 in our example.  

Cluster state will switch to "UPDATING", but all resources will be UP on both nodes.

You can track scaling progress and activities from "Work Requests" page under Resources section in VM cluster details home page.

Once scaling successfully completes, cluster's state will changed back to "AVAILABLE". Under "Resource allocation" section, Exadata Storage (TB) will show the new total usable storage.

7. Connect to any of the VM cluster and run "asmcmd lsdg" command to check ASM disk groups new size.

ExaCC: Scale VM Cluster Local Storage Size

 

- Overview:

• With multiple VMs per Exadata Infrastructure feature release, you can scale up or down the following resources on VM cluster:
• Local storage /u02. Scaling memory requires a rolling restart of the cluster nodes to take effect.
• Memory. Scaling memory requires a rolling restart of the cluster nodes to take effect.
• OCPU. Scaling OPCU is Online.
• Exadata storage. Scaling OPCU is Online.  
• You specify how much space is provisioned from local storage to each VM. 
• This space is mounted at location /u02, and is used primarily for Oracle Database homes. 
• The minimum supported size of /u02 is 60 GB per VM.
• The maximum supported size of /u02 is 900 GB per VM. Depending on the number of other VMs in the system and the space consumed by those VMs.
• A rolling restart of each VM is required to resize /u02.

In this blog, I'll demonstrate the steps to scale up local file system size using OCI console.

Steps to Scale Up VM Cluster Local File System Size

1.  Sign in to your OCI tenancy where your Exadata Database Service on Cloud @ Customer system is deployed.

2. Navigate to "Oracle Database" > "Oracle Exadata Database Service on Cloud@Customer" > "Exadata VM Clusters".

3. Click on your VM cluster name.

4. On VM cluster details home page, click "Scale VM Resources" button.

  

5. On "Scale VM Cluster" screen, under "Resource allocation per VM" section, enter local file system size per VM then click "Save Changes" button.

 

You can track scaling progress and activities from "Work Requests" page under Resources section in VM cluster details home page.

Once scaling successfully completes, cluster's state will changed back to "AVAILABLE". Under "Resource allocation" section, Local Storage (GB) will show the new total size for the cluster.

Also you can confirm new size from VM node by running df -h command.

 

ExaCC: Patch VM Cluster Grid Infrastructure

 

- Overview:

• Oracle is responsible to perform patches and updates to EXACC infrastructure every quarter (there is no patch in Dec). That includes:
• Physical compute nodes (Dom0).
• Network switches.
• Power distribution units (PDUs).
• Integrated lights-out management (ILOM).
• Exadata storage servers.
• The customer is responsible for patching Oracle GI Clusterware.
• Patching Oracle GI will be applied in the rolling manner to preserve database availability throughout the patching process.
• Patching prerequisites
• /u01 on the database host file system has at least 15 GB of free space.
• The Oracle Clusterware is up and running on both VM cluster’s nodes.
• The VM cluster’s nodes are up and running.
• Patching tools
• Oracle OCI console.
• DBAASCLI command line utility from one of VM cluster nodes.

In this blog, I'll demonstrate the steps to patch Grid Infrastructure on ExaCC VM cluster to the latest available release update (19.21 Oct/2023 at the time of writing this blog) using OCI console.

To apply a patch to Oracle Grid Infrastructure using DBAASCLI command line utility from one of VM cluster nodes. A high level steps:

1. Connect to VM as opc user.

2. Download GI patch from OCI control Plane server.

sudo dbaascli cswlib download --product GRID --version 19.21.0.0.0

sudo dbaascli cswlib showImages --product grid

3. Run Precheck.

sudo dbaascli grid patch --targetVersion 19.21.0.0.0 --executePrereqs

4. Apply patch.

sudo dbaascli grid patch --targetVersion 19.21.0.0.0 

Steps to Patch Grid Infrastructure on VM Cluster

1.  Sign in to your OCI tenancy where your Exadata Database Service on Cloud @ Customer system is deployed.

2. Navigate to "Oracle Database" > "Oracle Exadata Database Service on Cloud@Customer" > "Exadata VM Clusters".

3. Click on your VM cluster name.

4. On VM cluster details home page, next to "Updates Available" click "View updates" link.

5. Run patching Precheck. 

If the precheck is not successful, then the patch is not applied, and the system displays a message that the patch cannot be applied because the precheck failed.

    - On Patches releases page, you will a list of available release update patches for "Virtual Machine OS Update" and "Grid Infrastructure Release Update".

    - Review the list of available GI updates. Click the Actions icon (three dots) at the end of the row listing the patch you are interested in, and then click "Precheck" action, then run "Run Precheck" button in confirm precheck screen.

Note: once Precheck finishes, update state will be "Available". On "Update History" page, update state will be "Precheck Passed".

6. Apply GI Update.

Note: Patching of GI on cluster nodes will be executed in the rolling mode to preserve database availability throughout the update process.

Click the Actions icon (three dots) at the end of the row listing the patch that passed Precheck, and then click "Apply Grid Infrastructure Patch" action, then run "Apply Patch" button in confirm screen.

You can track patching progress from "Work Requests" page under Resources section in VM cluster details home page.

Note: once patching finishes, on "Update History" page, update state will be "Applied".

 

On VM cluster details page, Grid Infrastructure Version will be updated.

Congratulations, patching GI on ExaCC cluster nodes is completed !!!. 

ExaCC: Patch VM Cluster OS Image

 

- Overview:

• Oracle is responsible to perform patches and updates to EXACC infrastructure every quarter (there is no patch in Dec). That includes:
• Physical compute nodes (Dom0).
• Network switches.
• Power distribution units (PDUs).
• Integrated lights-out management (ILOM).
• Exadata storage servers.
• The customer is responsible for managing patches and updates of OS for VM cluster’s nodes.
• The customer can apply Oracle Exadata software release updates to VM cluster’s nodes at his/her convenience.
• Patching of VM cluster’s nodes will be executed in the rolling mode to preserve database availability throughout the update process.
• VM cluster patching tools:
• Oracle OCI console.
• patchmgr utility. The customer can run the utility either from one of ExaCC VM cluster nodes or from another server running Oracle Linux.

In this blog, I'll demonstrate the steps to patch/update ExaCC VM cluster OS image to the latest available release update using OCI console.

Steps to Update ExaCC VM cluster OS Image

1.  Sign in to your OCI tenancy where your Exadata Database Service on Cloud @ Customer system is deployed.

2. Navigate to "Oracle Database" > "Oracle Exadata Database Service on Cloud@Customer" > "Exadata VM Clusters".

3. Click on your VM cluster name.

4. On VM cluster details home page, next to "Updates Available" click "View updates" link.

5. Run patching Precheck. 

If the precheck is not successful, then the patch is not applied, and the system displays a message that the patch cannot be applied because the precheck failed.

    - On Patches releases page, you will a list of available release update patches for "Virtual Machine OS Update" and "Grid Infrastructure Release Update".

    - Review the list of available software updates. Click the Actions icon (three dots) at the end of the row listing the patch you are interested in, and then click "Precheck" action, then run "Run Precheck" button in confirm precheck screen.

Note: once Precheck finishes, update state will be "Available". On "Update History" page, update state will be "Precheck Passed".

6. Apply Exadata OS Image Update.

Note: Patching of VM cluster’s nodes will be executed in the rolling mode to preserve database availability throughout the update process.

Click the Actions icon (three dots) at the end of the row listing the patch that passed Precheck, and then click "Apply Exadata OS Image Update" action, then run "Apply Exadata Image Update" button in confirm screen.

Note: once patching finishes, on "Update History" page, update state will be "Applied".

On VM cluster details page, Exadata Image Version will be updated.

7. Confirm Exadata image version by running "imageinfo" command on VM cluster nodes.

Congratulations, patching VM cluster OS is completed !!!.  

ExaCC: Adding SSH Keys to Exadata Cloud @ Customer

 

- Overview:

• You can connect to the virtual machines in an Exadata Database Service on Cloud @ Customer system by using a Secure Shell (SSH) connection.
• An SSH private key file that corresponds to a public key that is registered in the system.
• When you create a VM cluster on your Exadata Database Service on Cloud @ Customer system, you must specify the public key portion of one or more SSH key pairs.
• You can also register extra keys separately after you create the VM cluster. 
• The public keys are stored in the authorized_keys file at ~/.ssh/authorized_keys. Separate authorized_keys files are located under the home directories of the operating system users. 
• By default, only the opc user account has an authorized_keys entry and is able to log in remotely. Do not remove or alter the automatically generated entry in authorized_keys for the opc user.

Steps to Add SSH Keys to ExaCC VM Cluster

1. Generate SSH key pairs.

    - Connect to SSH terminal and generate keys using command "ssh-keygen -t rsa". 

       The command will generate private and public keys.

    - Copy key pairs files to your desktop.

2. Sign in to your OCI tenancy where your Exadata Database Service on Cloud @ Customer system is deployed.

3. Navigate to "Oracle Database" > "Oracle Exadata Database Service on Cloud@Customer" > "Exadata VM Clusters".

4. Click on your VM cluster name.

5. On VM cluster details page, click "Add SSH Keys" button. 

6. On "Add SSH Keys" page, select "Upload SSH key files", browse and select public key file generated in the first step, then click "Save Changes" button.

Note: Cluster status will be changed to "UPDATING" for around 10 minutes. The cluster will stay available (no downtime).

7. Test SSH to VM cluster. 

     - Connect to SSH terminal

     - Run SSH command "ssh -i <SSH private key file> opc@<ExaCC VM cluster node IP>"

You are connected now to opc user on the cluster node !!!.