Remotely restarting a logical partition

You can restart an AIX®, Linux®, or IBM® i logical partition remotely by using the Hardware Management Console (HMC) command-line interface. You can run up to four concurrent simplified remote restart operations for a destination server. When the HMC is at version 8.5.0 or later, you can run up to 32 concurrent simplified remote restart operations for a destination server.

Before you begin

Before you begin, complete the following tasks:

You must ensure that the HMC is at version 8.2.0 or later, and the firmware is at level FW820, or later, and the VIOS must be at version 2.2.3.4 with the VIOS interim fix IV63331m4a or later, for the simplified remote restart feature.
You must ensure that the source and the destination servers are connected to the same HMC. However, when the HMC is at version 8.5.0 or later, the source and destination server can be connected to a different Hardware Management Consoles. In this case, the HMC at the source and destination servers must be at version 8.5.0, or later.
When the HMC is at version 8.3.0, or later, you can restart a logical partition on another server only when the source server is in the Initializing, Power Off, Error, or Error - Dump in progress state and the destination server is in the Operating state. When the HMC is at version 8.4.0, or later, you can restart a logical partition on another server only when the source server is in the Initializing, Power Off, Powering Off, Error, or Error - Dump in progress in progress state and the destination server is in the Operating state. When the HMC is at version 8.5.0, or later, you can restart a logical partition on another server only when the source server is in the Initializing, Power Off, Powering Off, No Connection, Error, or Error - Dump in progress state, and the destination server is in the Operating state.
When the HMC is at Version 9.2.950, or later, and the firmware is at a level FW950, or later, for the remote restart operation to be successful on the logical partition that supports platform keystore capability, the user-defined system key that is configured on both the source and the destination system must match.

Procedure

To remotely restart a logical partition, type the following command from the HMC command line:
```
 rrstartlpar -o restart -m <source managed system> -t <target manged system> -p <lpar name> 
```
When the source server and the destination server are managed by different Hardware Management Consoles, type the following command to remotely restart a logical partition:
```
rrstartlpar -m <source managed system> -t <target managed system> -p <lpar name> | --id <lpar id> 
--ip <IP address> [-u <user ID>] -o restart
```
Note: If a logical partition that is enabled with the simplified version of the remote restart feature was in the Suspended state before the simplified remote restart operation started, the simplified remote restart operation fails. You can use the --force option to force a simplified remote restart operation.

To verify the status of the simplified remote restart operation, type the following command from the HMC command line:
```
lssyscfg -r lpar -m <server> -F name,state,remote_restart_status
```
You can also run the lsrrstartlpar command to view the status of the simplified remote restart operation.
To specify the destination shared processor pool for the simplified remote restart operation, type the following command from the HMC command line:
```
rrstartlpar -m <source managed system> -t <target managed system> -p <lpar name>
| --id <lpar id> --ip <IP address> [-u <user ID>] - i|f 
“shared_proc_pool_name=<spp name>|shared_proc_pool_id=<spp id>” -o restart
```
Where:
- shared_proc_pool_name is the name of the shared processor pool on the destination server.
- shared_proc_pool_id is the ID of the shared processor pool.
Note: The shared_proc_pool_id and shared_proc_pool_name attributes are mutually exclusive.
To specify the destination virtual Fibre Channel mapping for the simplified remote restart operation, type the following command from the HMC command line:
```
rrstartlpar -m <source managed system> -t <target managed system> -p <lpar name> | --id <lpar id> 
--ip <IP address> [-u <user ID>] - i|f “virtual_fc_mappings=
slot_num/vios_lpar_name/vios_lpar_id/[vios_slot_num]/[vios_fc_port_name]” -o restart
```
Where:
- slot num is virtual Fibre Channel slot number.
- vios_slot_num is virtual Fibre Channel slot number of the VIOS.
To remotely restart a partition when the source server is in the No Connection state, type the following command from the HMC command line:
```
rrstartlpar -m <source managed system> -t <target managed system> -p <lpar name> | --id
<lpar id> -o restart --noconnection
```
When the HMC is at version 8.5.0, or later, an auto cleanup operation on the server is performed by the HMC after the successful completion of the simplified remote restart operation. The source server is back to the Standby or Operating state, and the Resource Monitoring and Control (RMC) connection for the Virtual I/O Server (VIOS) partitions that are serving the remotely restarted logical partitions becomes active.
You can enable the auto cleanup operation by typing the following command from the HMC command line:
```
rrstartlpar -o set -r <source managed system> -i "auto_cleanup_enabled=1" 
```
You can disable the auto cleanup operation by typing the following command from the HMC command line:
```
rrstartlpar -o set -r <source managed system> -i "auto_cleanup_enabled=0" 
```
The auto cleanup operation is enabled by default. If the remote restarted partitions are not cleaned up automatically, type the following command from the HMC command line to perform a cleanup operation on the source server:
```
rrstartlpar -o cleanup -m source managed system -p lpar name 
```
The default behavior of a cleanup operation is to remove the reserved storage device from the device pool on the source server. You can use the --retaindev option when you want to retain the reserved storage device in the pool and override the default action of the cleanup operation.
To remotely restart a logical partition with a different number of dedicated processors or virtual processors on the destination server than what the partition was assigned on the source server, or with a different number of shared processing units on the destination server than what the logical partition was assigned on the source server, type the following command from the HMC command line:
```
rrstartlpar -o restart -m <managed system> -t <target system> [-p <partition name>
| --id <partition ID>] -i “desired_procs = <desired procs absolute value>|min, desired_proc_units =
<desired proc units absolute value>|min”
```
The value for the desired_procs attribute indicates the number of processors with which the logical partition can be restarted. The following are the possible values for this attribute:
- desired procs absolute value - Must be greater than or equal to the current minimum dedicated processors or virtual processors, and less than or equal to current maximum dedicated processors or virtual processors.
- desired proc units absolute value - Must be greater than or equal to the current minimum processing units, and less than or equal to current maximum processing units.
- min - The value of the current minimum processing units, or the current minimum dedicated processors or virtual processors. In the case of dedicated processors, the logical partition is started with the minimum number of dedicated processors or virtual processors the logical partition was assigned on the source server. In the case of shared processors, the logical partition is started with the minimum number of shared processing units the partition was assigned on the source server.
To remotely restart a logical partition when you want to start the logical partition with a different amount of memory on the destination server than what the logical partition was assigned on the source server, type the following command from the HMC command line:
```
rrstartlpar -o restart -m <managed system> -t <target system> [-p <partition name> |
 --id <partition ID>] -i “desired_mem = <desired absolute memory>|min”
```
The value for the desired_mem attribute indicates the amount of memory with which the logical partition can be restarted. The following are the possible values for this attribute:
- desired_absolute_memory - Must be greater than or equal to the current minimum memory, and less than or equal to current maximum memory.
- min - The value of the current minimum memory. The logical partition is started with the minimum amount of memory the logical partition was assigned on the source server.
To remotely restart a logical partition when you want to start the logical partition with a different virtual switch on the destination server than what the logical partition was assigned on the source server, type the following command from the HMC command line:
```
rrstartlpar -o restart -m <managed system> -t <target system> [-p <partition name> | 
--id <partition ID>] -i “vswitch_mappings = <vlanId_1>/<source_vswitchName_1>/
<target_vswitchName_1>, …, <vlanId_n>/<source_vswitchName_n>/<target_vswitchName_n>”]”
```
When you want to prevent a logical partition from being started during the simplified remote restart operation, type the following command from the HMC command line:
```
rrstartlpar -o restart -m <managed system> -t <target system> [-p <partition name>
 | --id <partition ID>] --skippoweron
```
You can test the simplified remote restart operation of a shutdown partition from the source server, which is in Operating or Standby state by typing the following command from the HMC command line:
```
rrstartlpar -o restart -m <managed system> -t <target system> [-p <partition name>
| --id <partition ID>] --test
```