This section explains how to change the settings for the admin server.
Use the following procedure to change the IP address used on the admin LAN by the admin server.
Log on to the admin server as the administrator.
Stop the manager.
For details, refer to "5.1 Manager".
Change the SNMP trap destination set for the management blade and LAN switch blades.
Set the SNMP trap destination to the new IP address of the admin server.
Note
Depending on the LAN switch blade used, setting an SNMP trap destination may restrict SNMP access to that switch blade.
In a clustered manager configuration, set the physical IP addresses of both the primary and secondary nodes as SNMP trap destinations.
If the LAN switch blade is set to only grant access from known IP addresses, be sure to give permissions to the physical IP addresses of both the primary and secondary cluster nodes, as is done with trap destination settings.
For details, refer to the manual of the LAN switch blade to be used.
Change the IP address set within the operating system.
Change the IP address following the instructions given in the operating system's manual.
If the admin LAN has been made redundant, change the admin IP address set in the following tools or products.
PRIMECLUSTER GLS
BACS
Intel PROSet
Refer to the manual of each product for usage details.
In a clustered manager configuration, change the cluster IP address according to the instructions given in "Changing the IP Address of a Clustered Manager".
Change the IP address registered as the manager's admin IP address.
Use the "rcxadm mgrctl modify" command to set a new IP address.
[Windows]
>"Installation_folder\Manager\bin\rcxadm" mgrctl modify -ip IP_address <RETURN> |
[Linux]
# /opt/FJSVrcvmr/bin/rcxadm mgrctl modify -ip IP_address <RETURN> |
In a clustered manager configuration, for details on how to change the admin IP address registered for the manager, refer to "Registering the Admin IP Address of a Clustered Manager".
Log in to the admin server with an OS administrator account.
Change ServerView Agent settings on the managed server.
Change the SNMP trap destination of the ServerView Agent. Refer to the ServerView Agent manual for details on changing SNMP trap settings.
Stop the agents on managed servers. [Windows/Hyper-V] [Linux/Xen]
For details on how to stop agents, refer to "5.2 Agent".
Change Agent settings. [Windows/Hyper-V] [Linux/Xen]
Use the "rcxadm mgrctl modify" command to set the new manager IP address.
[Windows/Hyper-V]
>"Installation_folder\Agent\bin\rcxadm" agtctl modify -manager IP_address <RETURN> |
[Linux/Xen]
# /opt/FJSVrcxat/bin/rcxadm agtctl modify -manager IP_address <RETURN> |
Restart the agents on managed servers. [Windows/Hyper-V] [Linux/Xen]
For details on how to restart agents, refer to "5.2 Agent".
Repeat steps 6 to 11 for each managed server on which an agent is running.
Restart the manager.
For details on how to restart the manager, refer to "5.1 Manager".
Re-configure the HBA address rename setup service.
When using the HBA address rename function, change the IP address of the admin server that is set for the HBA address rename setup service according to "6.2.2.1 Settings for the HBA address rename setup service".
Back up managed servers.
If system image backups have already been collected, it is recommended to update those images in order to reflect the changes made above. For details on system image backup, refer to "8.2 Backing Up System Images" in the "ServerView Resource Coordinator VE Operation Guide".
System images backed up before changing the admin IP address of the admin server cannot be restored anymore after the change. It is recommended to delete all system images collected before change, unless those images are specifically needed.
Re-collect cloning images. [Windows/Linux]
If cloning images have already been collected, it is recommended to update those images to reflect the changed made above. For details on cloning image collection, refer to "8.2 Collecting a Cloning Image".
Cloning images collected before changing the admin IP address of the admin server cannot be deployed anymore after the change. It is recommended to delete all cloning images collected before change, unless those images are specifically needed.
Changing the IP Address of a Clustered Manager
In a clustered manager configuration, use the following procedure to change the IP address set within the operating system.
[Windows]
Change the IP address using the [Failover Cluster Management] dialog.
[Linux]
Stop the manager's cluster service.
Stop the manager's cluster service from the cluster administration view (Cluster Admin).
Log in to the admin server's primary node.
Log in to the operating system of the admin server's primary node with administration privileges.
Mount the shared disk on the primary node.
Mount the admin server's shared disk on the primary node.
Change the logical IP address.
Release PRIMECLUSTER GLS virtual interface settings from the PRIMECLUSTER resource, then change the PRIMECLUSTER GLS configuration.
For details, refer to the PRIMECLUSTER Global Link Services manual.
Activate the logical IP address.
Use the PRIMECLUSTER GLS command-line to activate the logical IP address.
For details, refer to the PRIMECLUSTER Global Link Services manual.
Registering the Admin IP Address of a Clustered Manager
In a clustered manager configuration, use the following procedure to register an IP address as the manager's admin LAN IP address.
[Windows]
Cancel registry replication settings.
On the primary node, bring online the shared disk and IP address, and take all other resources offline.
Next, remove the following registry key from the registry replication settings set for the "PXE Services" cluster resource.
For x64 architectures
SOFTWARE\Wow6432Node\Fujitsu\SystemcastWizard\DHCP
For x86 architectures
SOFTWARE\Fujitsu\SystemcastWizard\DHCP
Use the following procedure to remove the registry key.
In the [Failover Cluster Management] window, right click the "PXE Services" resource in "Summary of RC-Manager"-"Other Resources", and select [Properties] from the popup menu.
The [PXE Services Properties] dialog is displayed.
In the [Registry Replication] tab, select the above registry key and click <Delete>.
The selected key is removed from the [Root registry key] list.
After removing the registry key, click <Apply>.
Once settings are applied, click <OK> to close the dialog.
Change the manager IP address on the primary node.
On the primary node, use the "rcxadm mgrctl modify" command to set the new IP address.
>"Installation_folder\Manager\bin\rcxadm" mgrctl modify -ip IP_address <RETURN> |
Restore registry replication settings
Restore the registry key deleted in step 1 to the registry replication settings of the "PXE Services" resource.
Use the following procedure to restore the registry key.
In the [Failover Cluster Management] window, right click the "PXE Services" resource in "Summary of RC-Manager"-"Other Resources", and select [Properties] from the popup menu.
The [PXE Services Properties] dialog is displayed.
In the [Registry Replication] tab, click <Add>.
The [Registry Key] dialog is displayed.
Restore the registry key deleted in step 1 in "Root registry key" and click <OK>.
After completing registry key settings, click <Apply>.
Once settings are applied, click <OK> to close the dialog.
Assign the manager's shared disk and IP address to the secondary node.
In the Failover Cluster Management tree, right-click [Services and Applications]-[RC-manager], and select [Move this service or application to another node]-[1 - Move to node node_name].
The name of the secondary node is displayed in node_name.
Change the manager IP address on the secondary node.
On the secondary node, use the "rcxadm mgrctl modify" command to set the new IP address.
Use the same IP address as the one set in step 2.
Assign the manager's shared disk and IP address to the primary node.
In the Failover Cluster Management tree, right-click [Services and Applications]-[RC-manager], and select [Move this service or application to another node]-[1 - Move to node node_name].
The name of the primary node is displayed in node_name.
On the primary node, take the manager's shared disk and IP address offline.
[Linux]
Change the IP address set for the admin LAN.
Set a new IP address on the primary node using the following command.
# /opt/FJSVrcvmr/bin/rcxadm mgrctl modify -ip IP_address <RETURN> |
De-activate the admin server's logical IP address.
Use the PRIMECLUSTER GLS command-line interface to de-activate the logical IP address.
For details, refer to the PRIMECLUSTER Global Link Services manual.
Register the logical IP address as a PRIMECLUSTER resource.
Use the PRIMECLUSTER GLS command-line interface to register the virtual interface as a PRIMECLUSTER resource.
For details, refer to the PRIMECLUSTER Global Link Services manual.
Un-mount the shared disk.
Un-mount the shared disk from the primary node.
Log in to the admin server's secondary node.
Log in to the operating system of the admin server's secondary node with administration privileges.
Change the logical IP address.
Use the PRIMECLUSTER GLS command-line interface to remove virtual interface settings from the PRIMECLUSTER resource, register the resource and change the PRIMECLUSTER GLS configuration.
For details, refer to the PRIMECLUSTER Global Link Services manual.
Change the cluster configuration.
Use the cluster RMS Wizard to change the GLS resource set in the cluster service of either one of the cluster nodes.
After completing the configuration, save it and execute the following operations.
Configuration-Generate
Configuration-Activate
Start the Manager's cluster service.
Use the cluster administration view (Cluster Admin) to start the Manager's cluster service.
Resource Coordinator VE requires the following services to be running. When starting these services, make sure that the ports they are using do not conflict with the ports used by other applications or services. If necessary, change the ports used by Resource Coordinator VE services.
[Windows]
Manager services
Resource Coordinator Manager
Resource Coordinator Task Manager
Resource Coordinator Web Server (Apache)
Resource Coordinator Sub Web Server (Mongrel)
Resource Coordinator Sub Web Server (Mongrel2)
Systemwalker MpWksttr
Deployment services
Deployment Service
TFTP Service
PXE Services
[Linux]
Manager services
rcxmanager
rcxtaskmgr
rcxmongrel1
rcxmongrel2
rcxhttpd
mpwksttr-trap
Deployment services
scwdepsvd
scwpxesvd
scwtftpd
Change the ports used by the above services if there is a possibility that they will conflict with other applications or services.
For Windows operating systems, an ephemeral port may conflict with a Resource Coordinator VE service if the maximum value allowed for ephemeral ports (5000 by default) was changed. In this case, change the port numbers used by Resource Coordinator VE services to values greater than the maximum ephemeral port value.
This section explains how to change the ports used by Resource Coordinator VE services.
For information on how to change the ports used by ServerView Operations Manager, refer to the ServerView Operations Manager manual. The ports used for SNMP communication and server power control are defined by standard protocols and fixed at the hardware level, thus cannot be changed.
Refer to "Appendix C Port List", for details on the ports used by Resource Coordinator VE.
When using a firewall on the network, firewall settings should be updated to match the new port definitions and allow communications for any modified port.
Note
[Linux]
When the OS of the admin server is Linux, for detail about the changing procedure of the port which uses mpwksttr-trap, refer to the "Release Notes for the ServerView Trap Server" attached to ServerView Trap Server for Linux (trpsrvd).
The following table shows the product references used for this document.
Before replacement | After replacement |
|---|---|
Systemwalker Resource Coordinator | ServerView Resource Coordinator VE |
Systemwalker Resource Coordinator V12.0L30 or later | ServerView Resource Coordinator VE V2.1.1 or later |
Systemwalker Resource Coordinator Domain manager | ServerView Resource Coordinator VE Manager |
For information on ServerView Trap Server for Linux, refer to "2.1.1.3 SNMP Trap Daemon [Linux]" of the "ServerView Resource Coordinator VE Installation Guide".
Manager services
Use the following procedure to change the admin server ports used by manager services:
Stop manager services.
For details, refer to "5.1 Manager".
Change the port numbers.
Use the "rcxadm mgrctl modify" command to set a new port number for a given service name.
[Windows]
>"Installation_folder\Manager\bin\rcxadm" mgrctl modify -port name=number <RETURN> |
[Linux]
# /opt/FJSVrcvmr/bin/rcxadm mgrctl modify -port name=number <RETURN> |
In a clustered manager configuration, bring offline all manager resources except for the shared disk and IP address, move all cluster resources from the primary node to the secondary node, then execute the "rcxadm mgrctl modify" command on all the nodes that are hosting cluster resources.
Restart manager services.
For details on how to restart the manager, refer to "5.1 Manager".
Note
When changing the "rcxweb" port, the following ports should be set to the same value.
Admin client
Enter the "rcxweb" port in the Web browser URL used to log into Resource Coordinator VE.
If this URL is bookmarked in the Web browser "Favorites", change the port set in the bookmark's URL.
HBA address rename setup service
If the HBA address rename setup service is running, change the port number used to communicate with the admin server to the "rcxweb" port according to "6.3.3 Changing Settings for the HBA Address Rename Setup Service".
[Windows]
Change the RC console shortcut on the manager
Open the following folder on the admin server.
Installation_folder\Manager
Right click the "RC Console" icon, and select [Properties] from the popup menu.
In the [Web Document] tab, change the port number set in the "URL" field (as shown below).
URL: https://localhost:23461/ |
Click <OK>.
When changing the "nfagent" port, the following ports on managed servers should be set to the same value.
Set the "nfagent" port set on each managed server to the same value, according to the instructions given in "6.3.2.6 Changing Port Numbers".
The system image and cloning images collected before the change can no longer be used, and should be deleted.
If necessary, re-collect system images and cloning images.
Deployment services
Use the following procedure to change the ports used by deployment services.
[Windows]
Change the port numbers.
Open the Windows Registry Editor, and search for the following subkey:
When using a 32-bit version of Windows:
Key name: HKEY_LOCAL_MACHINE\SOFTWARE\Fujitsu\SystemcastWizard\CLONE
When using a 64-bit version of Windows:
Key name: HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Fujitsu\SystemcastWizard\CLONE
Select "PortBase" from the registry entries under this subkey.
From the menu, select [Edit]-[Modify].
The "Edit DWORD Value" dialog is displayed.
Select "Decimal" and click <OK>.
This port value will define the first port of the range used by deployment services.
However, because deployment services can use up to 16 port numbers, ensure that all ports included between "PortBase" (defined here) and "PortBase+15" do not conflict with any other applications or services. Moreover, be sure to set a value lower than 65519 for "PortBase" so that the highest port number ("PortBase+15") does not exceed the largest valid port number (65534).
In a clustered manager configuration, change port numbers on both the primary and secondary node.
Restart the server on which the port number has been changed.
[Linux]
Change the port numbers.
Edit the following file: /etc/opt/FJSVscw-common/scwconf.reg.
In PortBase (under HKEY_LOCAL_MACHINE\SOFTWARE\Fujitsu\SystemcastWizard\CLONE ), set the first value of the port number to be used by deployment services. This value should be entered in hexadecimal format. To avoid conflicts with ephemeral ports, use a value not included in the ephemeral port range defined by "net.ipv4 local port range".
This ensures that deployment services will use ports outside of the range defined by net.ipv4.ip_local_port_range for image operations.
Moreover, be sure to set a value lower than 65519 for "PortBase" so that the highest port number ("PortBase+15") does not exceed the largest valid port number (65534).
In a clustered manager configuration, change port numbers on both the primary and secondary node.
Restart the server on which the port number has been changed.
Information
Deployment services allow managed servers to boot from the network using a dedicated module stored on the admin server during backup, restore or cloning.
Note that changing port numbers on the admin server alone is enough to support communication during the above image operations. Therefore, no additional configuration is required on the managed servers.
Use the following procedure to change the maximum number of system image versions.
Change the maximum number of system image versions.
Confirm the maximum number of system image versions.
For details, refer to "5.4 rcxadm imagemgr" of the "ServerView Resource Coordinator VE Command Reference".
Note
If the specified limit is smaller than the number of existing system image versions, older versions will not be deleted automatically. In this case, backing up a new system image, will only delete the oldest version.
Delete unused image versions manually if they are no longer necessary. For details, refer to "8.5 Deleting a System Image" in the "ServerView Resource Coordinator VE Operation Guide".
If the RC console was already opened, refresh the Web browser after changing the maximum number of system image versions.
Use the following procedure to change the maximum number of cloning image versions.
Change the maximum number of cloning image versions.
Confirm the maximum number of cloning image versions.
For details, refer to "5.4 rcxadm imagemgr" of the "ServerView Resource Coordinator VE Command Reference".
Note
If the specified limit is smaller than the number of existing cloning image versions, older versions will not be deleted automatically. In this case, collecting a new cloning image version will require selecting a previous image version for deletion.
Delete unused image versions manually if they are no longer necessary. For details, refer to "8.5 Deleting a Cloning Image".
If the RC console was already opened, refresh the Web browser after changing the maximum number of cloning image versions.
Use the following procedure to change the location (path) of the image files folder.
Select the [Image List] tab in the RC console and confirm the current image list.
Log on to the admin server as the administrator.
Stop manager services.
For details, refer to "5.1 Manager".
Change the location of the image file storage folder.
Change the location of the image file storage folder according to the instructions given in "5.4 rcxadm imagemgr" of the "ServerView Resource Coordinator VE Command Reference".
Because image files are actually copied over to the new location, this step may take some time to complete.
In a clustered manager configuration, for details on how to change the image folder location, refer to "Clustered Manager Configuration".
Start manager services.
For details on how to restart the manager, refer to "5.1 Manager".
Select the [Image List] tab in the RC console and confirm the image list is same as before.
Clustered Manager Configuration
Settings differ depending on the operating system used for the manager.
[Windows]
Cancel registry replication settings.
Bring the shared disk and IP address online, and take all other resources offline.
Next, remove the following registry key from the registry replication settings set for the "Deployment Services" cluster resource.
For x64 architectures
SOFTWARE\Wow6432Node\Fujitsu\SystemcastWizard\ResourceDepot
For x86 architectures
SOFTWARE\Fujitsu\SystemcastWizard\ResourceDepot
Use the following procedure to remove the registry key.
In the [Failover Cluster Management] window, right click the "Deployment Service" resource in "Summary of RC-Manager"-"Other Resources", and select [Properties] from the popup menu.
The [Deployment Service Properties] dialog is displayed.
In the [Registry Replication] tab, select the above registry key and click <Delete>.
The selected key is removed from the [Root registry key] list.
After removing the registry key, click <Apply>.
Once settings are applied, click <OK> to close the dialog.
Change the location of the image file storage folder.
Change the location of the image file storage folder according to the instructions given in "5.4 rcxadm imagemgr" of the "ServerView Resource Coordinator VE Command Reference".
Because image files are actually copied over to the new location, this step may take some time to complete.
Run the "rcxadm imagemgr" command from either node of the cluster resource.
The new location should be a folder on the shared disk.
Restore registry replication settings.
Restore the registry key deleted in step 1 to the registry replication settings of the "Deployment Service" resource.
Use the following procedure to restore the registry key.
In the [Failover Cluster Management] window, right click the "Deployment Service" resource in "Summary of RC-Manager"-"Other Resources", and select [Properties] from the popup menu.
The [Deployment Service Properties] dialog is displayed.
In the [Registry Replication] tab, click <Add>.
The [Registry Key] dialog is displayed.
Restore the registry key deleted in step 1 in "Root registry key" and click <OK>.
After completing registry key settings, click <Apply>.
Once settings are applied, click <OK> to close the dialog.
[Linux]
Mount the shared disk on the primary node.
Log in to the primary node with OS administrator privileges and mount the admin server's shared disk.
Change the location of the image file storage directory.
Change the location of the image file storage directory according to the instructions given in "5.4 rcxadm imagemgr" of the "ServerView Resource Coordinator VE Command Reference".
Because image files are actually copied over to the new location, this step may take some time to complete.
Run the "rcxadm imagemgr" command on the primary node.
The new location should be a directory on the shared disk.
Un-mount the shared disk from the primary node.
Un-mount the shared disk (for which settings were performed in step 2) from the primary node.
