Home > Veritas Storage Foundation™ Volume Manager Manual Pages
VXMEND (1M) |
Maintenance Commands |
vxmend [-fprsvV] [-g diskgroup] [-o useopt] [-U usetype] fix how name [arg...]
vxmend [-fprsvV] [-g diskgroup] [-o useopt] [-U usetype] off name...
vxmend [-fprsvV] [-g diskgroup] [-o useopt] [-U usetype] on name...
Each invocation can be applied to only one disk group at a time. Any name operands will be used as record names to determine a default disk group, according to the standard disk group selection rules described in vxintro(1M). A specific disk group can be forced with -g diskgroup.
The field operand is a comma-separated list of keywords specifying fields to be cleared. Each keyword in the field operand is one of the following:
- all
- Clears all clearable utility fields. Normally, this clears all of the persistent and non-persistent utility fields. A usage-type utility may choose a different set of fields, as appropriate.
- putil
- Clears all persistent utility fields.
- putil0, putil1, or putil2
- Clears a specific persistent utility field.
- tutil
- Clears all non-persistent (temporary) utility fields.
- tutil0, tutil1, or tutil2
- Clears a specific non-persistent utility field.
Note: This particular functionality in conjunction with -f and -r flag can be used to recursively clear tutil0 fields in objects undergoing relayout, provided the top most object undergoing relayout is used to initiate it.
Usage types may implement additional field keywords. A usage type may also limit the set of clear operations that can be performed.
vxplex att v1 v1-01
is aborted by pressing the INTERRUPT key (or equivalent) several times (which prevents a clean abort of the operation), you may have to clear the operation with:
vxmend clear tutil0 v1-01 vxmend clear att1 v1 vxplex dis v1-01
Interruption of the vxplex att command will display a list of commands to run, which will include the previous list.
The number of plexes currently being attached, either directly or as part of a compound operation, is stored in the tutil0 field of the volume record as a string in the form ATTnumber. If the number drops to zero, the tutil0 field is cleared. Some operations on a volume require that the tutil0 field be cleared.
Similar to the above example, if an online relayout operation, either started using vxassist or the vxrelayout command is aborted by pressing the the INTERRUPT key (or equivalent) several times (which prevents a clean abort of the operation), you may have to clear the operation with:
vxmend -rf clear tutil0 topmost_volume | topmost_plex
Interruption of an online relayout operation will display a command similar to above with the appropriate object name to clear the tutil0 field. Online relayout stores a string RELAYOUT in the the tutil0 of all objects subject to relayout. If an interruption caused a cleanup to happen correctly, the tutil0 of all objects used by online relayout will be cleared, else they may or may not be cleared. The vxmend utility of the form mentioned above can be used to clear the tutil0 field. It is important that the object name used in the operation is the topmost object name or any object above the topmost object undergoing relayout.
- vxmend fix active plex
- Sets the state for the named plex to ACTIVE. The state for the volume is set to SYNC. The associated volume must be disabled, and the named plex must be in the STALE state.
When starting a volume in the SYNC state, all ACTIVE plexes are enabled and are synchronized to have the same contents using a special read/write-back recovery mode. Any STALE plexes are then recovered by copying data from the ACTIVE plexes.
- vxmend fix clean plex
- Sets the state for the named plex to CLEAN. The associated volume must be disabled, the named plex must be in the STALE state, and the volume must have no additional plexes in the CLEAN state.
A volume is not startable if one plex is in the CLEAN state and some plexes are in the ACTIVE state. Thus, several vxmend fix operations are normally used in conjunction to set all plexes in a volume to STALE and then to set one plex to CLEAN. A volume start operation will then enable the CLEAN plex and recover the STALE plexes by copying data from the one CLEAN plex.
- vxmend fix empty volume
- Sets the named volume and all of its associated plexes to the EMPTY state. The volume can then be re-initialized using vxvol start or any of the vxvol init operations. This operation requires that the volume be disabled.
- vxmend fix stale plex
- Sets the state for the named plex to STALE. The associated volume must be disabled, and the named plex must be ACTIVE or CLEAN. This operation names plexes that will be recovered by copying data from other plexes by a vxvol start operation.
Taking the last enabled read-write plex in a volume offline, or the last complete plex (if there are additional sparse plexes), normally fails unless -f is specified. Applying this operation to a volume also disables the volume, and does not require use of -f.
If a volume is disabled, then the checks for the last plex or the last complete plex are performed on the set of ACTIVE and CLEAN plexes, rather than on enabled read-write plexes.
An offline state for a plex can be cleared with vxmend on. Alternatively, an offline plex can be reattached with vxplex att.
- vxmend fix unstale subdisk
- Clears any flags indicating that a subdisk of a RAID-5 plex is invalid. This prevents the data on the subdisk from being recovered when the volume is started.
See vxintro(1M) for a list of standard exit codes.
Last updated: 17 Jul 2008
Copyright ©2009 Symantec Corporation
All rights reserved.