mmvdisk pdisk command
Manages mmvdisk recovery group pdisks for IBM Spectrum Scale RAID.
Synopsis
mmvdisk pdisk list --recovery-group {all | RgName[,RgName]}
[--declustered-array DaName | --pdisk PdiskName]
[--replace | --not-ok | --ssd-endurance-percentage N] [-Y]
ormmvdisk pdisk list -L
--recovery-group {{all | RgNaame} [--declusterebd-array DaName | --pdisk PdiskName]
[--replace | --not-ok | --ssd-endurance-percentage N][-Y]
ormmvdisk pdisk replace --prepare
--recovery-group RgName
{--pdisk PdiskName | --location LocationCode}
[--force-rg] [--force]
or
mmvdisk pdisk replace --recovery-group RgName
{--pdisk PdiskName | --location LocationCode}
[--force-rg] [--force-fru] [-v {yes | no}]
or
mmvdisk pdisk replace --cancel --recovery-group RgName
{--pdisk PdiskName | --location LocationCode}
[--force-rg]
ormmvdisk pdisk change --recovery-group RgName
--pdisk PdiskName
[--identify {on | off}]
[--clear-error-counters]
[--diagnose |
--suspend |
--resume |
--revive |
--revive-failing |
--revive-slow |
--begin-service-drain |
--end-service-drain |
--simulate-dead |
--simulate-failing]
Availability
Available on all IBM Spectrum Scale editions.
Description
Use the mmvdisk pdisk list command to show the pdisk information for selected pdisks. Pdisks can be selected for listing based on recovery group, declustered array, or pdisk maintenance state. The --not-ok flag will list pdisks that are not in service. The --replace flag will list pdisks that are marked for replacement. The --ssd-endurance-percentage N flag will list SSD pdisks with write endurance percentage greater than or equal to N.
By default, the mmvdisk pdisk list command prints a one-line summary for each listed pdisk. If the -L flag is used, pdisks are listed in stanza format, which shows detailed pdisk attributes and is quite voluminous.
Use the mmvdisk pdisk replace command to perform pdisk replacement when an IBM Spectrum Scale RAID recovery group is reported to need service.
- Run the mmvdisk pdisk replace --prepare command to prepare the pdisk for physical removal.
- Physically remove the disk and replace it with a new disk of the same type.
- Run the mmvdisk pdisk replace command to complete the replacement.
The mmvdisk pdisk replace --cancel command can be used to cancel the replacement of a pdisk that has been prepared for replacement. Canceling a prepared replacement does not render a required replacement unnecessary.
The mmvdisk pdisk change command changes the runtime state of a pdisk. Incorrect use of this command can cause file system or recovery group data to become unavailable or even unrecoverable. It should only be used in extreme situations under the guidance of IBM Service.
Parameters
- mmvdisk pdisk list
- List pdisks.
- mmvdisk pdisk replace
- Replace failed pdisks.
- mmvdisk pdisk change
- Change pdisk states. This is a low-level command that should only be used in extreme situations under the guidance of IBM service.
- --recovery-group all | RgName[,RgName...]
- With the mmvdisk pdisk list command, specifies the recovery groups from which to list pdisks.
- --declustered-array DaName
- Restricts pdisk listing to the specified declustered array.
- --pdisk PdiskName
- Specifies a pdisk by name.
- --replace
- Specifies that only pdisks that are marked for replacement are to be listed.
- --not-ok
- Specifies that only pdisks that are not functioning correctly are to be listed.
- --ssd-endurance-percentage N
- Specifies that only SSD pdisks with write endurance percentage greater than or equal to N are to be listed. Valid values for N are whole numbers from 0 to 100.
- --recovery-group RgName
- With the mmvdisk pdisk replace and mmvdisk pdisk change commands, specifies the recovery group for the target pdisk.
- --prepare
- With the mmvdisk pdisk replace command, prepares a pdisk for removal and replacement.
- --cancel
- With the mmvdisk pdisk replace command, cancels the preparation of a pdisk for removal and replacement.
- --location LocationCode
- With the mmvdisk pdisk replace command, specifies a pdisk by location code. Pdisk location codes can be found in the long form stanza output of the mmvdisk pdisk list command.
- --force
- With the mmvdisk pdisk replace --prepare command, prepares a working pdisk that is not marked for replacement. The --force flag does not permit the replacement of an otherwise good pdisk. It merely suspends the pdisk from use for temporary removal, inspection, and reseating of the same exact physical disk; a different disk must not be inserted. To resume operation of a good pdisk that was suspended with --force, only the mmvdisk pdisk replace --cancel command is permitted.
- --force-rg
- With the mmvdisk pdisk replace command, indicates that the pdisk being replaced is in a multi-disk carrier that also contain pdisks that are not in a recovery group, or in a different recovery group.
- --force-fru
- With the mmvdisk pdisk replace command, permits the replacement disk to have a different FRU (or type) than the removed pdisk.
- -v {yes | no}
- With the mmvdisk pdisk replace command, specifies whether the replacement disk should be verified to be empty of IBM Spectrum Scale RAID data before it is incorporated into the recovery group. The default is yes, verify that the pdisk is empty of data.
- --identify {on | off}
- Specifies that the identify light for the pdisk should be turned on or off.
- --clear-error-counters
- Specifies that the pdisk error counters for the pdisk should be cleared.
- --diagnose
- Runs basic tests on the pdisk. If no problems are found, the pdisk state automatically returns to ok.
- --suspend
- Suspends I/O to the pdisk until a subsequent --resume command is
given. If a pdisk remains in the suspended state for longer than a predefined timeout period,
IBM Spectrum
Scale RAID begins rebuilding the data from that pdisk
into spare space. Attention: This option is to be used with caution and only when instructed to perform disk maintenance manually, bypassing the automatic system provided by the mmvdisk pdisk replace command.
If a pdisk is removed using this option, vdisks that store data on the removed pdisk will become temporarily degraded, requiring data that was stored on the removed pdisk to be rebuilt from redundancy. Also, if you try to remove more pdisks than the redundancy level of the least redundant vdisk in that declustered array, data will become inaccessible. Therefore, when preparing to remove a pdisk, use the --begin-service-drain and --end-service-drain options instead of this option.
- --resume
- Cancels a previously given --suspend command and resumes use of the
pdisk.
Use this option only when instructed to perform disk maintenance manually, bypassing the automatic system provided by the mmvdisk pdisk replace command.
- --revive
- Attempts to make an out-of-service pdisk usable again by removing dead, failing, and readonly pdisk state flags. Data allocated on the disk that has not been rebuilt onto spare space can become readable again; however, any data that has already been reported as lost cannot be recovered.
- --revive-failing
- Like the --revive parameter, except that additional diagnostics particular to the failing pdisk state flag are attempted.
- --revive-slow
- Like the --revive parameter, except that additional diagnostics particular to the slow pdisk state flag are attempted.
- --begin-service-drain
- Starts draining the pdisk so that it can be temporarily removed. After issuing the command with
this option, wait until the pdisk has been completely drained before removing the pdisk. Note: This process requires that there be sufficient spare space in the declustered array for the data that is to be drained.
- --end-service-drain
Starts returning drained data to a pdisk after it has been brought back online.
- --simulate-dead
- Forces the failure of a pdisk by setting the simulatedDead pdisk state flag.
Attention: This option must be used with caution. If the total number of failures in a declustered array exceeds the fault tolerance of any vdisk in that array, permanent data loss might result.
- --simulate-failing
- Forces the failure of a pdisk by setting the simulatedFailing pdisk state flag.
Attention: This option must be used with caution. If the total number of failures in a declustered array exceeds the fault tolerance of any vdisk in that array, permanent data loss might result.
- -L
- Specifies that the mmvdisk pdisk list command produce long form stanza output.
- -Y
- Specifies that the mmvdisk pdisk list command produce colon-delimited raw output.
Exit status
- 0
- Successful completion.
- nonzero
- A failure occurred.
Security
You must have root authority to run the mmvdisk pdisk command.
The node on which the command is issued must be able to execute remote shell commands on any other node in the cluster without the use of a password and without producing any extraneous messages.
Example
To list the pdisks in recovery group ESS01L:
mmvdisk pdisk list --recovery-group ESS01L
The system displays output similar to the following:
declustered
recovery group pdisk array paths capacity free space FRU (type) state
-------------- ------------ ----------- ----- -------- ---------- --------------- -----
ESS01L e1d1s01 DA1 2 5588 GiB 216 GiB 38L6721 ok
ESS01L e1d1s02 DA1 2 5588 GiB 216 GiB 38L6721 ok
ESS01L e1d1s04 DA1 2 5588 GiB 216 GiB 38L6721 ok
ESS01L e1d1s05 DA1 2 5588 GiB 216 GiB 38L6721 ok
...
Example
To show all pdisks marked for replacement in all recovery groups:
mmvdisk pdisk list --replace --recovery-group all
The system displays output similar to the following:
recovery group pdisk priority FRU (type) location
-------------- ------------ -------- --------------- --------
ESS01L e3d4s06 5.78 38L6721 Enclosure SV50919775 Drawer 4 Slot 6
ESS01L e6d1s02 5.78 38L6721 Enclosure SV50918970 Drawer 1 Slot 2
mmvdisk: A lower priority value means a higher need for replacement.
Example
To prepare pdisk e6d1s02 in recovery group ESS01L for replacement:
mmvdisk pdisk replace --prepare --recovery-group ESS01L --pdisk e6d1s02
The system displays output similar to the following:
mmvdisk: Suspending pdisk e6d1s02 of RG ESS01L in location SV50918970-1-2.
mmvdisk: Location SV50918970-1-2 is Enclosure SV50918970 Drawer 1 Slot 2.
mmvdisk: Carrier released.
mmvdisk:
mmvdisk: - Remove carrier.
mmvdisk: - Replace disk in location SV50918970-1-2 with type '38L6721'.
mmvdisk: - Reinsert carrier.
mmvdisk: - Issue the following command:
mmvdisk:
mmvdisk: mmvdisk pdisk replace --recovery-group ESS01L --pdisk 'e6d1s02'
Example
To replace pdisk e6d1s02 of recovery group ESS01L after a new disk has been inserted:
mmvdisk pdisk replace --recovery-group ESS01L --pdisk e6d1s02
The system displays output similar to the following:
mmvdisk:
mmvdisk: Preparing a new pdisk for use may take many minutes.
mmvdisk:
mmvdisk: The following pdisks will be formatted on node ess01io1:
mmvdisk: /dev/sdrk
mmvdisk:
mmvdisk: Location SV50918970-1-2 is Enclosure SV50918970 Drawer 1 Slot 2.
mmvdisk: Pdisk e6d1s02 of RG ESS01L successfully replaced.
mmvdisk: Resuming pdisk e6d1s02#047 of RG ESS01L.
mmvdisk: Carrier resumed.