StorNext Connect Failovers
Note: This information applies to high-availability (HA), dual-node Quantum appliances only.
StorNext Connect is run on server Node 2 of the Quantum appliance. Server Node 1 will be operating as the primary node when the appliance is first brought online, and Node 2 will be operating as the secondary (failover/standby) node. This is the normal operating state of the appliance server nodes.
If a failover were to occur, causing both StorNext and StorNext Connect to be co-located on the same server node, the resources used by StorNext Connect could affect the performance of StorNext. A script, called snc_disable.py, is invoked by a Cron job once per minute that checks the status of snhamgr to determine the state of the server nodes. The script shuts down StorNext Connect when it detects that both the StorNext and StorNext Connect services are running on the same server node (which occurs after a failover). This allows StorNext to use all available necessary system resources.
Did you know?
During StorNext upgrades, the appliance fails over twice so that after the upgrade completes on both nodes, the server nodes are set back to the system state prior to the upgrade, with Node 1 operating as the primary. When Node 2 is operating as secondary again, StorNext Connect is enabled.
Mode |
Operation |
---|---|
Enabled |
In normal operation mode, server Node 1 is operating as primary and server Node 2 is operating as secondary, and StorNext Connect is left enabled. If the Node 2 is "Primary", StorNext Connect is set to run in disabled mode. |
Disabled |
When StorNext Connect detects that it is running as "not Primary," it will restore StorNext Connect operations, including the StorNext Connect interface. An admin user for StorNext Connect can enable StorNext Connect manually (see Enable StorNext Connect Manually After Failover). When this is done, the carbon cache is turned back on, enabling StorNext Connect to run normally even while the StorNext service is currently running on server Node 2. |
While Quantum recommends leaving the default behavior in place, the GUI provides a way to enable both the StorNext Connect and StorNext services while server Node 2 is operating as the "Primary" node for StorNext.
Keep the following notes in mind when determining whether to enable StorNext Connect while Node 2 is operating as "Primary":
- Quantum recommends using the default behavior on failover.
- If you enable StorNext Connect while Node 2 is operating as "Primary," using the steps below, there may be an impact to StorNext performance until the system fails back to using Node 1 as the node operating as "Primary" for StorNext.
- It is important to fail back to Node 1 so it is again operating as "Primary" after a failover. (See your Quantum appliance hardware documentation for further information.)
If a user is logged in as admin and the system has failed over with Node 2 now operating as "Primary," the StorNext Connect GUI will display an error message with a link labeled "Enable StorNext Connect." If a user is logged in as an account other than admin, a message indicating to contact an admin to enable StorNext Connect is displayed.
For admin users:
- Log in to StorNext Connect as admin.
- Click the Enable StorNext Connect link in the error message.
Clicking this link restarts the StorNext Connect service, and, after a short delay, allows the StorNext Connect GUI to display once again.