Data ONTAP enables you to register and monitor the status of UPS (Uninterruptible Power Supply) devices you are using with your storage system. In addition, you can configure the timing of certain Data ONTAP events when a power loss occurs.
You use the ups command to register UPS devices with Data ONTAP, and to review their status. For more information about the ups command, see the na_ups(1) man page.
Data ONTAP provides two configurable values to help you manage your storage system in case of a power outage:
The warningtime option specifies when Data ONTAP generates a warning SNMP trap, AutoSupport message and log message.
The default value of the warningtime option is 300 seconds (5 minutes).
The criticaltime option specifies when Data ONTAP generates another SNMP trap, AutoSupport message and log message, and then starts shutting down the storage system.
The default value of the criticaltime option is 60 seconds (1 minute).
When a power loss occurs, the UPS device begins supplying power to your storage system from its batteries. The UPS can only supply power as long as its batteries still have enough charge; the UPS is there to give you time to shut down your storage system cleanly.
The shutdown process is outlined in the following tasks.
Note
If you do not have AutoSupport enabled, the AutoSupport messages will not be generated.
For many environments, you can simply use the default values of five minutes for warningtime and one minute for criticaltime. However, you are advised to make sure that these values are set appropriately for your environment to avoid any data loss in case of a power outage. The warningtime value should give you enough time to do whatever manual processes you need to do prior to system shutdown, and criticaltime should provide enough time for the system to shut down cleanly.
Factors that can affect the optimal value of these options for your environment include:
If your UPS cannot support the default timing values, then your storage system will not be able to shut down cleanly.
If you have a large number of users, a large number of CIFS sessions, or any other workload factors that require a longer time to shut down, you need to increase the warning and critical time values to ensure that the system has sufficient time to shut down cleanly.
You may need to change the shutdown event timings to adhere to a protocol or requirement in place at your company.
If, after reviewing this documentation, you decide that you need to change these values, you can do so using the registry command.
Note
You are strongly advised to contact technical support before changing the shutdown event timing values.
Copyright © 1994-2005, Network Appliance, Inc. All rights
reserved.
Part No. 210-01077_B0
File: 7maint14.htm
Last updated: 05-Dec-2005 17:16:00