Configuring giveback involves the following tasks:
You can shorten the client service outage during giveback by enabling an option (cf.giveback.check.partner) that checks whether the failed node is already recovered and waiting for giveback--that is, waiting for disk reservation release--when you enter the cf giveback command at the console of the takeover node. If the failed node is waiting for giveback, giveback proceeds as normal; if not, giveback is canceled.
Note
This option works only with platforms booted from CompactFlash. To determine whether your storage system boots from CompactFlash, see your Hardware and Service guide.
To enable shorter giveback time, complete the following step.
| 1 |
options cf.giveback.check.partner {on|off}
Note |
This option specifies the number of minutes to delay an automatic giveback before terminating CIFS clients that have open files. During the delay, the system periodically sends notices to the affected clients. If you specify 0, CIFS clients are terminated immediately.
This option is used only if automatic giveback is On.
To set takeover delay time for CIFS clients, complete the following step.
| 1 |
options cf.giveback.auto.cifs.terminate.minutes minutes Valid values for minutes are 0 through 999. The default is 5 minutes. |
This option, when on, specifies that automatic giveback should immediately terminate long-running operations (dump/restore, vol verify, and so on) when initiating an automatic giveback. When this option is off, the automatic giveback is deferred until the long-running operations are complete. This option is used only if automatic giveback is On.
To set automatic giveback to terminate long-running processes, complete the following step.
| 1 |
options cf.giveback.auto.terminate.bigjobs {on|off}
The on argument enables this option. The off argument disables this option. This option is On by default. |
Copyright © 1994-2005, Network Appliance, Inc. All rights
reserved.
Part No. 210-01386_C0
File: ch5_mgn9.htm
Last updated: 21-Dec-2005 11:32:00