Managing the system attributes

Edit online

After you synchronize the KSYS cluster by using the ksysmgr sync ksyscluster command, the KSYS subsystem sets up the default system-wide persistent attributes. The KSYS subsystem uses these system-wide persistent attributes for activities such as automatic rediscovery of the resources, notification of critical events, removal of duplicate notification.

By default, the KSYS subsystem sets up the following system attributes:

deep_discovery

You can enable or disable the deep_discovery attribute at the system level through this tunable. The deep discovery process is similar to the discovery operation, except that the deep discovery process runs automatically every 24 hours. During the deep discovery process, the KSYS subsystem collects all information that is collected during the quick discovery process and information about the storage systems. You can update the time for the deep discovery process by using the auto_discovery_time attribute.

ksysmgr modify system deep_discovery=<enable|disable>

By default, the deep_discovery attribute is enabled. If you change it to disable, KSYS subsystem does not discover the environment automatically for any new or modified resources.

auto_discovery_time

Specifies the time for daily discovery at which the KSYS subsystem rediscovers the environment automatically for any new or modified resources. By default, the value of the auto_discovery_time attribute is 00:00, which means the KSYS subsystem discovers the resources and updates its database about the virtual machines every 24 hours at 00:00 hour.

The HMC and VIOS are also involved in the rediscovery process to update information about the virtual machines. Therefore, if your environment is large (for example, hundreds of logical partitions in the data center), you might want to set this to a time with minimal load on the HMC, VIOS, and the underlying I/O subsystems. Any configuration changes to the hosts, virtual machines, disks, and any other entities (for example, addition of new disks to a virtual machine) are captured when the rediscovery occurs. This attribute also specifies the time during which any changes in the configuration setting can be lost if a disaster occurs before the rediscovery.

quick_discovery

You can enable or disable the quick discovery property. The quick discover operation checks the states of existing VMs and does not consider configuration changes. By default, the quick_discovery attribute is enabled.

quick_discovery_interval

You can set time interval in minutes for quick discovery operation. By default, the quick_discovery_interval attribute is 60 minutes.

custom_script_timeout

Specifies the timeout duration for a custom script. If the custom script is unable to complete the execution within the configured timeout duration, the KSYS subsystem terminates the execution of the custom script and proceeds with the next execution. By default, the custom_script_timeout attribute is 0, which is considered as no timeout is set, and the system waits indefinitely for the script to run. For more information, see Configuring the timeout duration for custom scripts.

trace_file_size

Specifies the size in MB of the trace files the IBM.VMR daemon uses. By default, the trace_file_size attribute is not set and the size of the trace files is dynamically determined by RSCT.

ksys_spooling

Enables or disables trace spooling for the IBM.VMR daemon. By default, the ksys_spooling attribute is disabled. For more information on trace spooling, see Configuring trace spooling.

spool_dest_dir

Specifies the destination directory for trace spooling files for the IBM.VMR daemon when the ksys_spooling attribute is enabled. If you enable the ksys_spooling attribute and the spool_dest_dir attribute is not specified, then the spool_dest_dir attribute will get a default value as /tmp/ksys/rm.

spool_dir_max_size

Specifies the size in MB of the trace spooling area for the IBM.VMR daemon when the ksys_spooling attribute is enabled. By default, the spool_dir_max_size attribute is not set.

lose_vios_redundancy

Specifies to start the virtual machines in another site without the dual VIOS setup in the target host. By default, the lose_vios_redundancy attribute is set to no, which means that the dual VIOS setup is maintained during the disaster recovery operation.

If your currently active site, which has dual VIOS configuration, fails, and one of the VIOS in the target host of the backup site is not functioning, you might want to recover the virtual machines with only one VIOS on the backup site and continue the workloads that are running in the active site. In this case, you can set this attribute to yes. However, if the virtual machines are started with single VIOS configuration on the backup site, and later if you want to move the virtual machines back to the previous site that has a dual VIOS configuration, you must manually add the second VIOS into the configuration. For more information, see Configuring VIOS partitions for a dual setup.

auto_reverse_mirror

Resets the storage data mirroring after a site recovery operation.

Currently, this attribute is always set to no. You must resynchronize the storage data mirroring manually by using the ksysmgr resync command.

notification_level

Enables or disables the notification for different types of events. The notification_level attribute supports the following values:

low (default): Only critical error events are notified.
medium: Critical and warning error events are notified.
high: All events, which include informational events, are notified.
disable: None of the events are notified.

Note: The KSYS subsystem sends site event notifications without considering the value of the notification_level system attribute. For more information, see Notification for the KSYS events.

dup_event_processing

Reduces duplicate event notifications. The email and script notifications that are related to the duplicate events are also disabled. The dup_event_processingattribute can have the following values:

yes (default): Notifies about only those events that are not repeated in the last 24 hours.
no: Notifies all the messages.

replication_type

Specifies the storage replication mode across the sites. The replication_typeattribute can have the following values:

async(default): Specifies the asynchronous replication mode in which data is transferred across sites in predefined timed cycles or in delta sets
sync: Specifies the synchronous replication mode in which the storage subsystem acknowledges to the host that it has received and checked the data.
shared: Specifies a shared mode in which a single storage system is connected to multiple sites.
policybased: Specifies policy-based replication mode.

Note: This value is only supported for SVC storage. For more information, see Policy-based replication.; You must also specify the source and the target sites that are associated with the storage replication operation.

replicationpolicyname: Specifies the name of the replication policy. This attribute can be specified only when the replication_type attribute is set to policybased.

network_mapping

Enables or disables the network mapping policies for all hosts and host groups across the active site and the backup site.

network

Creates network mapping policy for all hosts and host groups. This attribute can be specified only when the network_mapping attribute is set to enable.

You can create either a VLAN mapping policy or a virtual switch mapping policy. This attribute must be specified as follows:

[network=<vlanmap|vswitchmap> sites=<siteA,siteB> 
      siteA=<#,[#,...]> siteB=<#,[#,...]>

For more information, see Configuring the network mapping policy.

vm_auto_discovery

Enables or disables whether newly created VMs and undiscovered VMs will be managed automatically. By default, the vm_auto_discovery attribute is enabled. For more information, see VM auto-discovery.

min_redundancy_paths

Sets the minimum number of redundant paths at the backup site allowed if the backup site disks have fewer paths than the home site disks.

By default, the min_redundancy_paths attribute is 0, which disables the minimum redundancy path feature and requires that backup site disks have the same number of paths as the home site disks.

If the redundancy path is enabled and the XSD version of the host is earlier than version 8.0, during the discovery operation, the ksysmgr commands displays a warning message specifying that the redundancy path is not supported.

#ksysmgr discover hg HG1
Running discovery on Host_group HG1, this may take few minutes...
.
.
.
Discovery has finished for HG1
WARNING:Since Xsd_version for below host(s) is less than 8.0, Redundancy paths not supported
XYamur-MNOP-LQT-ABCDEF
XYyana-MNOP-LQT-GHIJKL

The XSD version 8.0 requires VIOS version 3.1.3.10, or later.

hmc_ping_timer

Specifies in seconds the period between scheduled health checkups of the HMCs when the ping attribute is enabled. By default, the hmc_ping_timer attribute is 0, which indicates a default period of 10 seconds.

sa_ping_timer

Specifies in seconds the period between scheduled health checkups of the storage systems when the ping attribute is enabled. By default, the sa_ping_timer attribute is 0, which indicates a default period of 10 seconds.

cleanup_files_interval

Specifies in days the maximum period since last modification time of temporary VMR log files before they are removed. By default, the cleanup_files_interval attribute is 7.

memory_capacity

Specifies memory capacity percentage for flexible capacity policy at system level. For more information, see Configuring the flexible capacity policies.

cpu_capacity

Specifies CPU capacity percentage for flexible capacity policy at system level. For more information, see Configuring the flexible capacity policies.

ksys_lang

Specifies language setting for ksysmgr and VMRM. By default, the ksys_lang attribute is set to en_US. For more information, see Enabling locale.

connection_timeout

Specifies in seconds the connection timeout for KSYS REST API calls. By default, the connection_timeout attribute is 75.

storage_group_update

Enables or disables automatic performance of EMC SRDF rdfg storage updates. By default, the storage_group_update attribute is disabled. For more information, see EMC RDF group auto update.

hmc_port_number

Specifies the inbound application port for the HMCs. By default, the hmc_port_number attribute is 443. You can modify the HMC port value to 12443.

vpmem_override

Enables or disables Virtual Persistent Memory (vPMEM). By default, the vpmem_override attribute is disabled. For more information, see vPMEM override.

ping

You can enable or disable the ping attribute at the system level through this tunable. The KSYS subsystem performs scheduled health checkup on the HMC and storage systems. The KSYS subsystem informs you about the health state of the HMC and storage systems through the ping attribute. If the ping attribute is disabled, the KSYS subsystem does not provide the health state notification about the HMC and storage systems. Also, when the ping attribute is disabled, the HMC_UNREACHABLE event and the STG_UNREACABLE event are not triggered. To set the ping tunable at the system level, run the following command:

ksysmgr modify system ping=<enable|disable>

By default, the ping attribute is enabled.

optimise_put_storage

Enables or disables whether storage information is sent to the VIOS in a single call with all disks of all workgroups or separate calls for each workgroup.

Enabling optimise_put_storage in an environment with many workgroups improves performance. By default, the optimise_put_storage attribute is disabled.

Examples for querying system properties and status

To query the information or status related to system-wide attributes, use the following command:
```
ksysmgr query system [properties | status [monitor=<no|yes>]]
```

To query the existing values of the system attributes, run the following command:

ksysmgr query system properties

An output that is similar to the following example is displayed:

System-Wide Persistent Attributes
auto_discovery_time:         04:00 hours
lose_vios_redundancy:        no
auto_reverse_mirror:         yes
notification_level:          medium
dup_event_processing:        no
replication_type:            Asynchronous
network_mapping:             Enabled
vlanmap:                     Not currently set
vswitchmap:                  Not currently set
drvlanmap:                   Not currently set
drvswitchmap:                Not currently set
quick_discovery_interval:    60 minutes
quick_discovery:             enable
deep_discovery:              enable
vm_auto_discovery:           enable
custom_script_timeout:       none
trace_file_size:             not set
memory_capacity:             Priority Based Settings
                             low:100
                             medium:100
                             high:100
CPU_capacity:                Priority Based Settings
                             low:100
                             medium:100
                             high:100
ping:                        disable
min_redundancy_paths:        disable
optimise_put_storage:        disable
User Scripts for Site:       None

User Scripts for Host Group: None

User Scripts for VM:         None

To query the system status, run the following command:
```
ksysmgr query system status
```
The output might look similar to the following sample:
```
Discovery is in progress for Site Site1.
 Discovery is in progress for Site Site2.
 Please use "ksysmgr query system status monitor=yes " to track the progress of the operation.
```
If you specify the monitor=yes parameter, the KSYS monitors and displays all the current actions on the KSYS node. The actions include discovery, verification, movement, and recovery operations, which continue to progress even when the command is killed or when you close the command window. The monitor parameter also displays the encountered errors that are logged by the virtual machines, hosts, disk pairs, and disk groups. For example:
```
ksysmgr query system status monitor=yes
        Discovery is in progress for Site Site1.
        Discovery is in progress for Site Site2.
        Monitoring status...
Running discovery on entire site, this may take few minutes...
…
```
To query the system status and monitor all entities of the system, run the following command:
```
ksysmgr q system status monitor=yes
```
This system-wide command monitors all entities of the KSYS subsystem. This command displays all operations that are being run on each entity, and it also displays all operations that were previously run on each entity.

To modify the system-wide attributes, use the following command:

ksysmgr modify system <attribute>=<new_value>

Examples for modifying the system-wide attributes

To enable the KSYS subsystem to automatically rediscover the resources at 04:00 every day, run the following command:
```
ksysmgr modify system auto_discovery_time=04:00
```
To change the notification level of your system to receive notification for all critical errors and warnings for all events, run the following command:
```
ksysmgr modify system notification_level=medium
```
To change the duplicate event processing option to receive notification for all events, even if the events are duplicated, run the following command:
```
ksysmgr modify system dup_event_processing=no
```
To change the storage replication mode between two sites from asynchronous mode to synchronous mode, run the following command:
```
ksysmgr modify system replication_type=sync sites=SiteA,SiteB
```
To set the minimum redundancy paths, run the following command:
```
ksysmgr modify system min_redundancy_paths=<no. of paths>
```