Hardware Requirements for Deploying SNS
Operating System Disk
Disk Protection: RAID 1 (recommended).
Capacity: Minimum 150 GB or higher.
Storage Type:
SSD storage is preferred but not mandatory.
HDD is acceptable if SSD is unavailable.
b. SSD/NVMe for Elasticsearch and Other Container Storage
Recommended Storage Type:
NVMe is preferred.
SSD is an alternative.
15K SAS drives are acceptable but may cause slower bucket listings.
c. Swarm storage requirements
Drives must be unformatted.
Drives should appear as single paths (e.g., /dev/sdc).
CPU Requirements
CPU Type
Supported: x86/64-based architecture.
Not Supported: ARM or RISC processors.
Higher core counts are recommended for better performance.
Minimum CPU count:
Clock Speed
Minimum: 2.2 GHz per core.
Lower clock speeds are supported but may result in reduced performance.
Network Switch Specifications
Minimum Capability: The network switch must support a minimum of 10 Gbps or 25 Gbps Ethernet (10/25GbE).
Note: Multipath things need to add here
RAM Specifications
Minimum Requirement: 64 GB RAM.
Server Operating System Requirements for SNS Deployment
Operating System
Required OS: Ubuntu 22.04.
Partitioning Recommendation:
The operating system partitions and /var/lib should be mounted on a separate RAID-protected SSD set. this is required to safeguard the operating system in the event of a single drive failure.
Note: Mimimum space for /var/lib should added
Network Configuration
IP Address: Ensure the server is assigned a static IP address for reliable communication and consistent networking.
Storage Drives
Drive Format:
Storage drives must remain unformatted.
Path Visibility:
Drives should appear as single paths in the system (e.g., /dev/sdc, /dev/sdd).
Known Issues
UI Notification for Free Ports on Network Switches
Issue: If there are any unused ports on the network switches, the UI may display the following warning:
"Nics on <server> has issues!"
Impact: This is a false-positive alert and does not affect the functionality of the system.
Initial Server Access and Configuration
Logging into the Server
Ensure Server Readiness:
The server must be running Ubuntu 22.04 and should be powered on and accessible on the network.
Access via SSH:
Use SSH to log in to the server:
ssh <username>@<server_ip>
Switch to Root User:
After logging in, switch to the root user:
sudo su
Enter the password configured during the server installation when prompted.
Installing the SSA Debian Package
Installation Command
To install the SSA Debian package, execute the following command as the root user:
sudo su -c "bash <(wget -qO- https://apt.cloud.datacore.com/installssa/install-ssa.sh)" root
Installation is done via script.
There are a few prechecks to accept.
It adds the apt repository, then installs the latest periferyssa debian package as shown in the example below:The apt repository for the Swarm Software Appliance has now been added to this machine. The script will now install the application pending some environmental checks Do you want to proceed? (yes/no)
Accepting those prechecks adds the Swarm deb repo.
Type ‘yes’ and press enter to proceed with the installation of the bootstrapper and required packages. Once the installation is complete, a URL will be provided to access the UI. Following this, the support bundle packages will be updated, and firewall settings will be configured. Use the provided URL to access the UI.
Automatic reboots after unattended upgrades have been disabled. The bootstrapper and api services have now been installed you can now go to http://<serverIP>:9010/ui to continue configuration
If the installer is at “Scanning linux images…” stage, press return and it should proceed further.
Note: Before accessing the UI to create the cluster, please check if the template file has been generated. Run the command ls /etc/hosttype to verify the existence of the file. If the file exists, execute cat /etc/hosttype to check its contents. If the file is empty, contains the template name "test," or if the file is not generated, halt the cluster creation process and contact the CENG team to resolve the issue. Once the issue is fixed by the CENG team, you can proceed with the cluster creation as mentioned below. This will help avoid reinstalling the Debian package again.
Access the provided URL (http://<serverIP>:9010/ui) from any browser, and it will prompt you with a "Create Cluster" button on the page. Click the "Create Cluster" button to proceed.
It will open the "Create New Cluster" page. Follow the steps below:
Provide the desired cluster name.
Upload the license using the "Choose File" link.
If no license is uploaded, the default 2TB license will be applied.
Click on the "Advanced" dropdown link.
Provide the Domain Name and Bucket Name for creation during cluster setup.
If Domain Name and Bucket Name are not provided, the default Domain Name “test“ and Bucket Name “default” will be used.
After entering all details, click the "Next" button to proceed.
A confirmation page will open. Follow the steps below:
Verify that the entered details are correct.
If all values are correct, click "Confirm" to proceed further.
If any modifications are needed, click "Back" to modify the values, then click "Next" and "Confirm" again to proceed.
Wait for a few minutes while the provided details are validated and the initial configuration is completed. Once all checks are done, you will see a pop-up message saying "New cluster is created successfully." Click "Done" to finish.
Once the "Calculating..." window opens, click on "Show Details." This will display the current phase of the cluster creation process.
Wait for a minimum of 15 minutes, depending on network speed. After that, the "Log in to Single Node Swarm" page will open. Provide the default username and password as "periferyadmin/password" and click on login. Note on cases when the instalation happend on machine with multiple empty NIC card ports you will see an exclamation mark as shown below the same can be ignore and proceed.
After logging in, the "Update Password" page will open. Set the password as desired and click the "Update Password" button. Within a few seconds, the password will be updated, and a "Password Updated Successfully" message will briefly appear before disappearing. For the next login, use the updated password, as the initial default password will no longer work.
You Will Be Redirected to the Dashboard Page, and Now Your Cluster is Up and Running: See the Below Section in UI Which Will Navigate You Through
S3 Endpoint: In the "S3 Endpoint" section, view the S3 endpoint details.
Advanced Section: Under the "Advanced" section, you will see the "Content Management" and "Storage Management" link boxes, Clicking on these links will navigate you to the respective UIs in a new window.
Storage Box: Under the "Storage" box, observe the utilized and free space bars.
periferyadmin Dropdown Menu (located at the top-left corner of the UI):
About Single Node Swarm: Provides more details about "Single Node Swarm."
Change Password: Allows you to update your password.
Settings: Navigates to the "Access Control" page.
Shut Down: Shuts down the server.
Logout: Logs you out from the UI.
