(Cluster environments only) Provides an interface to perform various disk refit-related actions, such as showing the current status of the disk refit process, resetting the state of the disk refit process, skipping the disk refit process for an instance, and so on.
sp_refit_admin [‘help’] | 'status' | ['reset’ | 'skiperfit' [, instance_name]] |[ 'removedevice’, device_name]
displays information on sp_refit_admin syntax and usage.
displays the current status of the disk refit process. It lists all the instances and their private devices for which disk refit is still pending. If no such device exists, it prints a message saying so.
resets the state of the disk refit process. It takes an optional parameter instance_name.
If instance_name is not supplied, this parameter resets the disk refit process back to the beginning of Phase One, so that subsequent disk refit command starts the disk refit process from Phase One and refits all the regular shareable devices, as well as private devices of the instance.
If instance_name is supplied, this parameter resets the disk refit process back to the beginning of Phase Two for that instance, so that a subsequent disk refit command on that instance starts the disk refit process from Phase Two for that instance, and refits only the private devices of that instance.
skips running Phase Two of the disk refit process for one or all instances in the cluster without dropping the device. This parameter is meaningful only after the completion of Phase One of the disk refit process. It takes instance_name as an optional parameter.
removes a device from the disk refit process. This parameter requires the name of the device that is to be removed, as the input parameter device_name or instance_name.
Resets the state of the disk refit process to the start of Phase One:
sp_refit_admit 'reset'
After executing reset, the user must run Phase One and Phase Two of the disk refit process.
Resets the state of the disk refit process on the instance named 'cluster1_instance1' to the start of Phase Two for the instance:
sp_refit_admin 'reset', 'cluster1_instance1'
This interface removes sysdatabases entry for all the databases created on the private devices owned by 'cluster1_instance1', and the sysusages entries corresponding to the private devices owned by 'cluster1_instance1'. After executing, you must run Phase Two of disk refit on 'cluster1_instance1'.
Skips the disk refit process of all the refit-pending
private devices of instance 'cluster1_instance1':
sp_refit_admin 'skiprefit', 'cluster1_instance1'
This example removes the sysdatabases entry
for all the databases that use any of the refit-pending private
devices owned by 'cluster1_instance1',
and removes all the entries in sysusages for
all the deleted databases.
To skip the disk refit process on all the refit-pending private devices of all the instances in the cluster, enter:
sp_refit_admin 'skiprefit'
To remove the device ''device1' from the disk refit process:
sp_refit_admin 'removedevice', 'device1'
This action removes the sysdatabases entry
for all databases created on 'device1',
and all the sysusages entries corresponding to 'device1'.
It also removes 'device1' from sysdevices.
You must follow the instructions in Chapter 12, “Troubleshooting,” in the Clusters Users Guide after executing skiprefit, to ensure the consistency of the system tables before resuming normal operation.
Use removedevice only during the disk refit process, to remove the device from the refit process. Do not use it in place of sp_dropdevice
You can use sp_refit_admin even when the instance is started with the -m option and trace flag 3608 ON.
The permission checks for sp_refit_admin differ based on your granular permissions settings.
Granular permissions enabled |
With granular permissions enabled, you must be a user with manage disk privilege. |
Granular permissions disabled |
With granular permissions disabled, you must be a user with sa_role. |
For information on problems encountered with disk refit, see the Troubleshooting and Error Guide
