clvupdate Command

Purpose

Cleans up the system if the Live Update operation fails and if the default cleanup attempt is not successful.

Syntax

clvupdate [ -e ] [ -o ] [ -u ] [ -a HMChostname -U HMCusername] [-f AltCfFile]

clvupdate [ -e ] [ -n ] [ -u ] [ -a HMChostname -U HMCusername] [-f AltCfFile]

clvupdate -d

clvupdate -l

clvupdate -r

clvupdate -v

Description

You must clean up the system before you attempt to run a Live Update operation if the initial one failed and the default cleanup operation is not successful. The clvupdate command analyzes the state of the system and restores the system to its original state. An unsuccessful attempt during a default cleanup operation indicates that some condition is blocking the cleanup operation. The clvupdate command identifies the conditions that are blocking the cleanup operation and prompts the user with the necessary actions. After the user resolves all of the blocking conditions, the clvupdate command completes the cleanup operation.

The clvupdate command depends on the content of the /var/adm/ras/liveupdate/liveupdate.cf file that is generated during the failed Live Update operation. If the liveupdate.cf file is lost or corrupted, the clvupdate command cannot restore the state of the logical partition (LPAR). You must specify the previous liveupdate.cf file by using the -f option if you attempted another Live Update operation after the first operation failed.

It is recommended that you run the clvupdate command under the guidance of technical support if a failed Live Update operation caused a system failure or an unusable system state.

Parameters

Flags

Examples

1. If Live Update process fails and the Live Update lock is taken, run the clvupdate command with the -l option to remove the lock before running the clvupdate operation again. The following example shows the usage of the clvupdate command with the -l option on a display:

   clvupdate
   The Live Update lock is taken, indicating that a Live Update operation might be in progress.
   Use -l to forcefully remove the lock and proceed with cleanup anyway.
   clvupdate -l.
   clvupdate
   Cleanup action succeeded.

   Note: The Live Update operation keeps track of its state during the operation in the kernel. The Live Update operation uses the Live Update kernel state and Live Update kernel status kernel variables for coordinating operations that require strict synchronization. The clvupdate command checks these kernel variables before it proceeds with the cleanup operation. By checking these kernel variables, the clvupdate identifies situations in which a cleanup operation is not necessary. In some cases, you must specify additional options to bypass these checks, such as when the partition is rebooted after a failed Live Update operation.
2. In some cases, the Live Update kernel state might be clean but requires a cleanup. If the partition has been rebooted after the failed Live Update attempt, you must run the clvupdate command with the -e option.

   clvupdate
   The Live Update kernel state is INIT, which normally indicates that the system is in a clean state. 
   Use -e to proceed with cleanup anyway.
   clvupdate -e
   Cleanup action succeeded.

3. If the Live Update operation has aborted unexpectedly, the Live Update kernel status might be in a state that prevents the clvupdate command from running. To bypass this check, use the clvupdate command with the-u option.

   clvupdate
   The Live Update kernel status indicates that some processes from live update might still be running.
   Please run cleanup tool later or, if this is not the case, use -u to proceed with cleanup anyway.
   clvupdate -u
   Cleanup action succeeded.

4. If an administrator performed a manual cleanup operation but did not reset the Live Update kernel variables, you can run the clvupdate command with the -r option to reset the Live Update kernel variables to perform another Live Update operation. An example clvupdate command with the -r option follows:

   clvupdate -r

   Note: The Live Update operation uses the liveupdate.cf file to record important configuration information about the operation. The clvupdate command reads the liveupdate.cf file and identifies the operations that are needed to restore the partition to its original state. The liveupdate.cf file is located in the /var/adm/ras/liveupdate directory. The old liveupdate.cf file is renamed by appending a time stamp to its file name each time the Live Update operation is run. If the clvupdate command is run without specifying the -f option, it reads the default liveupdate.cf file.
5. If you want to perform a cleanup action to undo a change that was caused by a previous failed Live Update operation, use an alternative liveupdate.cf file instead of the default file. An example clvupdate command with the -f option follows:

   clvupdate -f liveupdate.cf.yyyy-mm-dd_HH:MM:SS.xxx
   Cleanup action succeeded.

6. An error can occur during a Live Update operation after the blackout time. When an error occurs past that point, the workload is moved to the surrogate partition, and the original partition is moved to an active state. You must run the clvupdate command with the -o option for the cleanup operation to proceed. An example clvupdate command with the -o option follows:

   clvupdate
   The clean up process has been aborted because the original partition is still active. 
   Turn off partition manually or use '-o' option to force shutdown first.
   clvupdate -o
   Cleanup action succeeded.

7. The Live Update operation has a framework that notifies other system components that are affected by the operation. The Live Update notification feature allows other system components to run scripts that coordinate with the Live Update operation at different phases. During one of the phases, you can run scripts to perform a cleanup operation that is specific to system components during a Live Update operation failure. The clvupdate command provides the option that runs only the scripts of other system components in certain scenarios. An example clvupdate command with the -n option on a display follows:

   clvupdate -n
   Cleanup action succeeded.

8. You can use an alternative HMC if the HMC that was used during the Live Update is unresponsive. You must authenticate to the HMC first. The following example shows the authentication process on a display:

   hmcauth -a hmc1 -u user1 -p password
   clvupdate -a hmc1 -U user1
   Cleanup action succeeded.

   Note: Surrogate boot disks are imported for debugging purposes if the Live Update operation in a PowerVC mode fails. In this case, the following message is displayed:

   Surrogate boot disk(s) have been imported as hdiskx for debugging purposes.
   Use clvupdate -d to remove them when done.

   If the disk is no longer needed, it can be deleted by running the clvupdate command with the -d option.
9. To delete the surrogate boot disk, enter the following command:

   clvupdate -d
   Cleanup action succeeded.

10. lvup_rootvg is the name of the volume group that the surrogate partition uses to start. The lvup_rootvg volume group is not needed after the partition is rebooted after a Live Update operation. The lvup_rootvg volume group is not automatically removed after a PowerVC mode Live Update operation. The clvupdate command provides an option to remove this volume group. You must authentic the partition with the PowerVC server to perform this operation because a PowerVC operation must be performed. The system displays an output that is similar to the following example:

    clvupdate -v
    1430-159 FAILED: No valid PowerVC session token. Run pvcauth.
    pvcauth -a pvchost -u pvcuser -u pvcpasswd
    clvupdate -v
    The temporary Live Update storage was successfully removed. Cleanup action succeeded.