Professional Documents
Culture Documents
Ref: CLUS40BE-1205
This documentation is provided "as is" without any warranty of any kind, either
expressed or implied, including (but not limited to) the implied warranties or
conditions of merchantability or fitness for a particular purpose. Atempo shall
not be liable for loss of profits, decreasing or interrupted business activity, for
loss of data or data use, nor for any indirect, special or consequential damages
whatsoever, even if Atempo has been advised of the possibility of such damag-
es arising out of a fault or an error in the documentation or in Time Navigator
software.
Atempo, Atempo’s logo and Time Navigator are registered trademarks of At-
empo.
All brand or product names mentioned in this guide are trademarks or regis-
tered trademarks of their respective companies or organizations.
Table of Contents
Introduction
Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Time Navigator Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
Time Navigator Release Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Time Navigator Getting Started Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Other Guides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3
Time Navigator Online Help . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Time Navigator Tutorials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4
Technical Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Web Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Generating the Environment Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5
Your Comments are Welcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Chapter 2 Installation
Installation on Windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Cluster Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11
Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Installation Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Step 1: Time Navigator Server Installation . . . . . . . . . . . . . . . . . . . . . . . . . 14
Step 2: Time Navigator for Cluster Agent Installation in Cluster Mode . . .14
Step 3: Time Navigator for Cluster Agent Installation in Local Mode . . . .16
Step 4: Disaster Recovery Solution Installation and Configuration . . . . . . . 17
Installation on Unix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18
Starting the Daemon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19
Installation on VMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20
Installation on NetWare . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Prerequistes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Agent Startup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26
ii Time Navigator for Cluster
Appendix
Appendix 1: FAQ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Appendix 2: Restoring Microsoft Cluster MSCS . . . . . . . . . . . . . . . . . . . . . . . . . . .45
Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Quorum Restore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Microsoft Technical Note Q245762 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Appendix 3: Restoring a Microsoft Cluster Node without a Time Navigator Disaster
Recovery Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Index
1
Introduction
Time Navigator is a data backup and archiving software. It also provides safe data restore
with the Time Navigation concept.
This manual contains information that will enable you to understand cluster principles, and
to install and manage Time Navigator for Cluster.
Conventions
This guide uses conventions to make information easy to access and understand.
■ Text in Time Navigator graphical application windows is displayed in the following
font:
Window text.
■ Time Navigator commands and scripts related to the different operating systems
(Unix, Windows, etc.) are displayed in the following font:
scripts and commands.
■ Cross-references to further information are indicated by the ☞symbol.
■ Procedures indicating the steps to follow to perform an action with Time Navigator
are presented as follows:
➤ Procedure title
1. First step.
2. Next step, etc.
■ Command syntax is presented as follows:
■ Square brackets [ ] indicate an option.
■ The - sign refers to a parameter. The parameter is either followed by the type
of information to provide, or is self-sufficient.
■ The | sign indicates a choice between several parameters.
Introduction 3
The PDF format requires the use of Adobe Acrobat Reader, which can be downloaded from
the Adobe web site, at www.adobe.com.
Note The Time Navigator Release Notes and Time Navigator Getting Started
Guide manuals are also available in print format.
The Time Navigator Getting Started Guide is an introductory guide to familiarize users
with Time Navigator. This guide contains information concerning Time Navigator
Server and Agent installation, backup configuration and restore methods.
Other Guides
The Specific guides describe installation, configuration and use of Time Navigator options,
such as Time Navigator Archiving Server, Time Navigator for Oracle, Time Navigator
for MS-SQL Server, Time Navigator for NDMP, etc.
4 Time Navigator for Cluster
This online help system allows you to find the information you need easily and quickly:
■ From any dialog box within the application, click on Help to access context-sensitive
information describing the parameters of the dialog box.
■ From any graphical interface, choose Help - Help Topics to get the online help for this
interface. You can also search for information through the Contents tab, the Index tab
or the Search tab that offers a powerful full-text search.
To access these tutorials, choose Help-Tutorial from the Time Navigator Administration
Console or Time Navigator Restore & Archive Manager. The tutorials are also
available from the Autorun (Windows) or from the CD-ROM index.htm file (Unix).
http://www.macromedia.com/shockwave/download/alternates/
Introduction 5
Note The Shockwave Player and Macromedia Flash Player programs are not
available for IBM AIX and VMS.
Technical Support
Web Support
If you have subscribed a maintenance contract and you encounter a problem with
Time Navigator, consult the Atempo Web Support at:
http://support.atempo.com
Complete information on Atempo Web Support, including how to get a login, the range of
tools and services available and how to use each of them, can be found in the
Troubleshooting Guide, on your Time Navigator delivery CD-ROM.
If you are unable to solve your problem or find the information you need, Atempo and its
partners will help you. Web support includes an interactive interface through which you can
log your inquiries directly and follow them up. Support is also available by email and by
phone.
When calling the Atempo Technical Support team, please provide your maintenance
contract number. We also recommend that for all Support inquiries you generate the
Time Navigator Environment Report as summarized in the next section, and explained
exhaustively in the Troubleshooting Guide.
The report is generated by the Time Navigator Environment Reporter which retrieves the
information that is essential to address your problems as quickly as possible:
■ Time Navigator version installed.
■ Operating system related information.
6 Time Navigator for Cluster
To automatically generate the environment report, run the Time Navigator Environment
Reporter as follows:
■ On Unix systems, run the $TINA_HOME/tina_env_report.sh script.
■ On Windows systems, the user must be part of the Administrator group and choose
Start-Programs-Time Navigator-Utilities-Time Navigator Environment Reporter.
A README file describing the operating mode is available at the following FTP address:
ftp://ftp.atempo.com/private/tina_env_report/README.txt
C HA PT ER 1
Cluster Overview 1
Definition
A cluster is defined as a group of two or more independent servers, known as nodes, linked
together in such a way that all servers in the cluster always contain the same information.
Should one of the servers fails, a second server immediately replaces it. The cluster appears
as a single entity on the network. Access to resources is managed by the cluster nodes via
one, or more, virtual machine.
To back up clusters, Time Navigator backs up the shared storage disk arrays via a virtual
machine (virtual IP address and its public network alias). The goal is to be able to back up
the same data from any of the cluster nodes whatever the resource controlling node is.
Warning Backing up local disks cannot be perfomed during a node switch because the
Time Navigator service is stopped and restarted on the other node.
Cluster Concept
With Windows NT, a cluster is made of one or more resource groups. Each group contains
at least:
■ A disk resource located in a SCSI or FC storage disk array,
■ A network resource (virtual IP address),
■ A network name.
8 Time Navigator for Cluster
Each node is connected to a storage disk array shared via one or more SCSI or FC buses. A
disk can only belong to a single node at any given time.
Nodes communicate with one another through a dedicated network called VPN (Virtual
Private Network) or "Heartbeat". If the heartbeat detects any problem with the active node,
it will switch the application to a different node.
The VPN also allows the cluster to know which application is currently active on which
node.
Users access cluster groups via virtual IP addresses and/or the network name.
Procedures
Switchover (failover and failback)
The administrator organizes cluster software resources in functional units called groups and
individually assigns them to each node by default.
Should an incident occur with the active node, the resources federated by a cluster group that
were running on the node are switched over to a different cluster node. Thus the user is
unaffected by the incident except for the restarting time and, eventually lost transactions.
The switching order is determined by the dependencies existing between the resources.
Chapter 1 Cluster Overview 9
3. The cluster restarts resources one by one on the second node, according to the
dependencies order. The resources are once more available.
Note Changing from a stable to an unstable state is a failover; the reverse is a failback.
☞ For the latest detailed list of supported version for each cluster, see the Time Navigator
Compatibility Guide on Atempo’s web site: www.atempo.com.
10 Time Navigator for Cluster
11
C HA PT ER 2
Installation 2
This chapter describes the Time Navigator for Cluster software installation on Windows,
Unix, VMS and NetWare. It also explains the cluster resources configuration procedures in
Time Navigator Administration Console.
The Time Navigator for Cluster installation and use require a good knowledge of the
cluster technology.
Installation on Windows
Cluster Components
When backing up a cluster, several components must be considered, each type being backed
up via different Time Navigator elements:
■ Shared storage resources
This component correspond to the virtual machine. The backup of the shared
resources is performed via Time Navigator for Cluster.
■ Node Local disks
They are backed up via a Time Navigator Agent for Windows for their file system
and a Time Navigator Disaster Recovery solution for their system disk: either
Time Navigator ASR for Windows 2003 and Windows XP, or Time Navigator
Disaster Recovery for Windows (Powered by WinPE).
Note If you are unable to use any Time Navigator Disaster Recovery solution, a
node restore procedure is described in “Appendix 3: Restoring a Microsoft
Cluster Node without a Time Navigator Disaster Recovery Solution”, page 48.
■ Quorum
The Quorum resource contains information pertaining to the cluster configuration, as
well as some log files, useful to administer the cluster.
12 Time Navigator for Cluster
The Quorum backup takes place via the Time Navigator System State application
located on the physical node hosting the Quorum. A System State application must be
declared on all cluster nodes.
☞ See the Time Navigator Agent for Windows Guide for details about the System
State application.
Prerequisites
User Permissions
To install Time Navigator, you must be logged in as administrator.
2. Within that group, right click on the disk component, select Properties and click on the
Parameters tab.
3. Make a note of the disk letters available for the installation of Time Navigator on the
virtual machine.
Installation Procedures
14 Time Navigator for Cluster
Note The installation procedures are described at length in the Time Navigator
Installation Guide. Only the points specific to the cluster installation are
discussed in this section.
Note Time Navigator, unlike Windows operating systems is case sensitive. Make sure
you respect the case of the host names when installing the cluster.
See the Time Navigator Installation Guide for specific instruction to install the server.
Note the communication port numbers used on the server as you need to use the same ones
for the agents installation.
4. Provide a user name and password to access the remote physical nodes (i.e., the nodes
other than the one from which the installation is taking place).
5. In the Installation Type screen, choose Agent.
6. In the Environment screen, you cannot choose to use the local node etc\services
file. Since the services file is located in an unshared directory,
(WINNT\system32\drivers\etc\) when the node resource switches, the
services file of the other node is not updated.
However, the services file is updated during installation in order to avoid conflicts
with other applications. When the service starts, the Time Navigator binaries use the
port numbers defined by the environment variables TINA_SERVICE_TCP_NUM and
TINA_SERVICE_UDP_NUM.
7. In the Installation Path screen, do not keep the default installation directory. Instead,
you must use a directory located on the active node, on one of the disk identified
earlier. See “Selection of the Installation Disk”, page 12 for details, in that example
the Installation path would be F:\Program Files\Atempo\tina..
8. Proceed with the remainder of the setup. See the Time Navigator Installation Guide
for details.
9. Repeat this procedure for all the cluster groups to back up.
hosts file
During the setup, a generic Time Navigator service is added to the Cluster Administrator,
in the virtual machine group.
Warning The Time Navigator service must imperatively be started with the Cluster
administrator, never with the Windows services manager. In the Windows
services manager, the Time Navigator service must be set to start manually.
To start the Time Navigator service, right click on the Time Navigator resource, in the
Cluster Administrator, and select Put on-line.
16 Time Navigator for Cluster
Installing Time Navigator on the local node is required for instance, to back up the
Quorum. The Quorum must be backed up via the Time Navigator System State application
Atempo recommends installing Time Navigator on every node of the cluster.
hosts file
Note The node name must be associated to an IP address on the public network. To
verify it, run a ping <hostname> command that will give you the associated
address. If there is no associated address, click on Start, Settings, Control Panel,
Network, Bind, All Protocols, TCP/IP Protocol.
Select the first network card and click on Down. You will be asked to restart the
system. This verification must be done for every node in the cluster.
Warning If the hosts.txt file is not updated, the Time Navigator resources will be
put offline on the cluster, interrupting all Time Navigator processes on the
virtual machines.
During the setup, the services file is updated and the Time Navigator service is set to
start automatically.
Note More than two networks may be associated to the nodes. For example: a LAN, a
VPN and an administration network to manage the backups.
If this is the case, make sure that the Time Navigator service created in the cluster
group has the right resources dependencies.
See the Time Navigator ASR for Windows 2003 and Windows XP, or Time Navigator
Disaster Recovery for Windows (Powered by WinPE) documentation, according to your
disaster recovery solution to install and configure it on the physical nodes of the cluster.
The installation procedure for a cluster node is identical to that of any agent.
18 Time Navigator for Cluster
Installation on Unix
To install Time Navigator on a Unix cluster, use the install.sh file that you find at
the root of the CD-ROM. You must perform the installation while being connected as the
root user.
When backing up a cluster, two types of components must be considered, each type being
backed up separately:
■ The shared storage resources (virtual machines)
■ The node local disks
As a result, Time Navigator Agent for Cluster must be installed on the shared storage
resource and, eventually on the physical nodes, if you want to back up the nodes’ local disks.
The installation procedure applies to both the virtual machine and the physical node
installation.
Note The installation procedures are described at length in the Time Navigator
Installation Guide. Only the points specific to the cluster installation are
discussed in this section.
Warning The Time Navigator installation must be done on the node where the resource
to back up is active.
1. Mount the Time Navigator CD-ROM and launch the installation program:
./install.sh -local_hostname <physical name or cluster alias
name>
2. Select the Time Navigator Agent installation.
3. You must use the environment variables to define the communication services. You
cannot refer to the local file /etc/services of the node because when the resource
switches, the file information is not updated on the other node.
4. Proceed with the remainder of the setup.
The installation program starts the daemon if all prerequisites are satisfied. However, you
must add the daemon startup in the resource startup and shutdown scripts. This must be done
with the system administrator.
1. Add the following lines at the end of the virtual resource startup script:
if you use a bash type shell,
#!/bin/sh
# Starting Time Navigator daemon
#
TINA_HOME=/<virtual_resource_path>/tina;export TINA_HOME
LD_LIBRARY_PATH=/<virtual_resource_path>/tina/Lib;export LD_LIBRARY_PATH
TINA_SERVICE_TCP_NUM=2525;export TINA_SERVICE_TCP_NUM
TINA_SERVICE_UDP_NUM=2526;export TINA_SERVICE_UDP_NUM
if [ -x /<virtual_resource_path>/tina/Bin/tina_daemon ]; then
/<virtual_resource_path>/tina/Bin/tina_daemon; echo "tina_daemon"
fi
if ( -x /<virtual_resource_path>/tina/Bin/tina_daemon ) then
/<virtual_resource_path>/tina/Bin/tina_daemon
echo "tina_daemon started"
else
echo "no tina_daemon binary"
endif
2. Add the following lines at the beginning of the virtual resource stopping script:
if you use a bash type shell,
#!/bin/sh
# Stopping Time Navigator daemon
#
TINA_HOME=/<virtual_resource_path>/tina;export TINA_HOME
LD_LIBRARY_PATH=/<virtual_resource_path>/tina/Lib;export LD_LIBRARY_PATH
TINA_SERVICE_TCP_NUM=2525;export TINA_SERVICE_TCP_NUM
TINA_SERVICE_UDP_NUM=2526;export TINA_SERVICE_UDP_NUM
if [ -x /<virtual_resource_path>/tina/Bin/tina_stop ]; then
/<virtual_resource_path>/tina/Bin/tina_stop; echo "tina_daemon stopped"
fi
if ( -x /<virtual_resource_path>/tina/Bin/tina_stop ) then
/<virtual_resource_path>/tina/Bin/tina_stop
echo "tina_daemon stopped"
else
echo "no tina_stop binary"
endif
Installation on VMS
The cluster installation is necessary only on one of the cluster nodes.
The HOSTS. configuration file must match the host name and the cluster alias. Do this by
adding the line: localhostname <cluster alias>.
It is necessary to start the Time Navigator service on all the cluster nodes.
Chapter 2 Installation 21
Installation on NetWare
To install Time Navigator on a NetWare cluster, you must manually copy the files. There
is currently no SMS installation module for NetWare 6x.
Note The files are the same as those used on physical nodes. The difference is in the
configuration of the Time Navigator Hosts.txt file.
Prerequistes
The cluster resources are localized by the SLP protocol. It is therefore important to ensure
that all servers (including virtual servers) be present in the list of SLP services (SMDR
section).
☞ If you encounter a problem with SLP, consult the Novell Knowledge Base through the link
(http://support.novell.com/search/kb_index.jsp).
22 Time Navigator for Cluster
2. From the Time Navigator CD-ROM, copy the NLM of the Bin directory (SMS
Client) into ncs_volume1:/Atempo/tinancs/Bin
3. From the Time Navigator CD-ROM, copy the mess.fc file in the Conf directory
into ncs_volume1:/Atempo/tinancs/Conf
4. Create the catalogs.txt, param.txt and hosts.txt files in ncs_volume1:/Atempo/tinancs/
Conf
5. Add Parameter:trunc_node_content=500 to the param.txt file.
6. Add localhostname VirtualServerName to the hosts.txt file.
7. Create the sys:system/tinancs.hom file on each node with the following content:
ncs_volume1:/Atempo/tinancs
Warning Make sure that you set a carriage return at the end of each line of all the
configuration files (tinancs.hom,catalogs.txt …).
Time Navigator in the Cluster environment must have its parameters set to avoid calls to
the etc/services file.
Chapter 2 Installation 23
The TCP and UDP ports are defined in the agent startup file. The loading and unloading
scripts of the virtual resources must be modified.
2. In the «$TINA_HOME/Conf» directory, create the Dntinancs.ncf file and define the
TCP and UPD ports.
For example:
load NCS_VOLUME1:/Atempo/tinancs/Bin/tina_stp.nlm -tina
tinancs -tina_service_tcp_num 25018 -tina_service_udp_num
25019
Chapter 2 Installation 25
3. Add the Time Navigator startup and shutdown files in the «$TINA_HOME»
directory.
4. Once the scripts are modified, place the resource offline and then online to take the
parameters into account.
26 Time Navigator for Cluster
Agent Startup
After load script modifications, the agent is automatically restarted. Simply loading a virtual
resource is sufficient to start the Time Navigator agent. However, it is sometimes
necessary to do this manually. This starts and stops the agent on the virtual resource.
In the case where the argument is /cluster=off, the clusterized volumes also appear while
viewing the node volumes containing the resource cluster while browsing.
The SMS application, which takes charge of the virtual server, is created as in the standard
mode:
Hosts + application
Backup
➤ To configure the Backup Class
1. Make sure that the classes are at level 4 in the structure or above:
In Case of Failover
In case of failover, the order of commands described in the load and unload scripts is used.
Time Navigator Agent for SMS unloads and reloads on a valid node. The IP reference of
the virtual server is once more operational and the backup continues after a restart timeout
on incident.
Note There is no apparent error on the SMS application. The backup continues from
the point where it was stopped.
Chapter 2 Installation 29
Logs Example
1. Here is an example of logs showing the events around a backup startup, a failover,
cancelled jobs, backup restarts and backup end:
30 Time Navigator for Cluster
Creating Hosts
You need to create hosts for all physical nodes and virtual machines.
➤ To create a host
1. In Time Navigator Administration Console, choose Platform-Create-Host.The host
creation window appears.
2. For a virtual resource, select the cluster type corresponding to your cluster in the Type
drop down list. For a physical host, choose the host type.
3. Enter the physical node or virtual resource name in the Host Name textbox. The host
name corresponds to the value entered in $TINA_HOME/Conf/hosts (Unix) or
%TINA_HOME\Conf\hosts.txt (Windows).
4. Click on OK to validate.
5. Right click on the host icon you have just created and select Enable to enable the host.
Tip If you have a lot of hosts to enable, it is recommended to use the Platform-Use-
Enable-Host menu because it allows for host multiple selection.
You need to create a System State application on every node in order to back up the Quorum.
☞ See the Time Navigator Agent for Windows Guide for details about the System State
application.
31
C HA PT ER 3
Cluster Management 3
This chapter presents the principles involved in managing clusters with Time Navigator.
Using a Cluster
☞ For more information on how to manage standard agents, see the chapter on Agents in the
Time Navigator Administration Guide.
Depending on your platform type (static or dynamic binaries), only one type of
Time Navigator application can be installed per group.
Note With Windows, some applications have cluster specific versions (example:
Exchange module cluster).
If a node fails (example: disk crash), all cluster resources, including the virtual IP address
and the mount points switch to the second node. The startup script of the virtual ressource
(including Time Navigator) is then run from this second node. Time Navigator can restart
backups of this node with the virtual IP address if the Retry on Incident option has been
configured.
32 Time Navigator for Cluster
☞ For more information on the Retry on Incident option, see the Creating a Strategy chapter
in the Time Navigator Administration Guide.
Chapter 3 Cluster Management 33
Backup
Method
■ To back up a cluster agent, you need to schedule one or several strategies associated
to one or several classes on the virtual resource directories.
■ To back up an application located on the virtual resource, select the host representing
the virtual resource in Time Navigator Administration Console (not one of the
nodes), choose Platform-Create-Application and define the required parameters for the
application.
Warning All the disk resources accessible by the virtual machine are listed in the User
interface. However you must not back up the disks belonging to a different
virtual machine nor the local node disks. This is due to the fact that when the
resource switches, only shared disks are switching.
Retry on Incident
When a resource is switched during backup, an error happens and the backup fails. You must
plan a retry on incident to allow backing up to start again. The delay before the retry takes
place must be greater than the resource switching time.
The switch of the group resource can be automatic (if a node overload is detected or
following a disk crash) or manual (depending on the administrator).
Windows Backup
To be able to restore all the elements of the cluster (quorum, disaster recovery of one node
or of the entire cluster) in addition to the data located on the share resources, you must
perform the following backups:
34 Time Navigator for Cluster
■ System State for each cluster node (see the Time Navigator Agent for Windows
guides for details)
■ Disaster Recovery for each cluster node (see the Time Navigator ASR for
Windows 2003 and Windows XP or the Time Navigator Disaster Recovery for
Windows guide according to the version of your operating system)
■ File system for each cluster node via the Time Navigator Agent for Windows,
excluding the system disk.
Chapter 3 Cluster Management 35
Restore
Method
The overall restore method is identical to the standard client restore. You have to select the
data to restore, synchronize and start the restore. However, if switching occurs, the restore
job aborts. You have to wait for the switch to be finished, restart restore deselecting the data
already restored.
☞ For more information on Restore, see to the Time Navigator Administration Guide and
the Time Navigator Restore Guide.
Note If you are unable to use any Time Navigator Disaster Recovery solution, a
node restore procedure is described in “Appendix 3: Restoring a Microsoft
Cluster Node without a Time Navigator Disaster Recovery Solution”, page 48.
■ File system backup of each node (excluding the system disk) using Time Navigator
Agent for Windows.
Note If all the other nodes are shut down when you perform the restore, the quorum will
be restored along with the node. If another node is active when you perform the
restore, the quorum will not be restored.
When you completely restore a cluster, two procedures are available: one without hardware
modification of the shared resource and a second one to use if the disk array hosting the
shared resources has changed since the backup.
1. Perform an ASR recovery of one of the nodes (see the Time Navigator ASR for
Windows 2003 and Windows XP guide for details).
2. Run the tina_load_clusdb process on the restored node to restore the quorum on
the shared resource.
3. Restore the node file system using the Time Navigator Restore & Archive
Manager.
4. Verify that the cluster is running.
5. Perform an ASR recovery of all the other nodes.
6. Verify that the cluster works properly when the active node switches.
This section applies to situations where you need to restore the quorum. Three cases are
described:
■ The disk hosting the quorum has crashed
■ The quorum is corrupted
■ You want to return to a previous version of the quorum
Prerequisites
In order to restore the cluster quorum, you must have access to the following elements:
■ System State backup of one of the nodes
38 Time Navigator for Cluster
■ Disk signature file updated after the cluster installation with the DUMPCFG tool of
the Windows 2003 Resource Kit.
Note This parameter allows the Cluster service to start on the node even though the
quorum is not available.
3. In the Cluster Administrator interface, right click on the cluster (at the root of the tree)
and select Properties.
4. Click on the Quorum tab and choose a different physical resource for the quorum.
5. Stop the Cluster service again on the node. Remove the start parameter and restart the
Cluster service.
6. Restart the Cluster service on all other nodes.
7. You must imperatively perform a new System State backup to make sure that it contains
the correct quorum configuration. Previous backups are valid only if the new quorum
disk has the same letter and the same signature as the pre-crash one.
8. If you want to assign the same letter and signature to the new quorum disk, proceed
as follows:
■ Start the Cluster service on every nodes.
■ Remove the resource corresponding to the crashed disk.
■ Replace the disk.
■ Restore the disk signature using the command :
DUMPCFG -S<signature><physical disk number>
■ Recreate the disk in the cluster.
■ Switch the quorum to that disk: in the Cluster Administrator interface, right
click on the cluster (at the root of the tree) and select Properties. Click on the
Quorum tab and choose the new disk for the quorum.
Time Navigator stops the service on the node and restores the cluster configuration
in the ClusDB file in the systemroot\cluster registry directory.
The node automatically restarts when the restore is complete.
4. Start the Cluster service on the other nodes.
Note This step forces the quorum log restore from the information contained in the
local node registry.
6. Use the ClusterRecovery tool located in the Microsoft Windows Resource Kit to
restore the additional quorum files.
7. Start the Cluster service on the other cluster nodes.
8. Remove the /resetquorumlog Starting Parameter on the first node.
☞ See the Time Navigator Agent for Windows documentation for details concerning
the System State Application.
Time Navigator stops the service on the node and restores the cluster configuration.
The cluster service restarts on that node and stops all the other cluster nodes.
Windows copies the restored quorum from the running node to the Cluster quorum
disk and to all other nodes.
3. Start the Cluster service on the other cluster nodes.
4. Verify that the cluster configuration is correct ant that it works properly including in
cases of active node switch.
5. Once the restore is complete, set the TINA_ACD_FORCE_CLUSTER environment
variable to No in the System State Application.
41
C HA PT ER 4
In addition to the Agent mode, you can also use Time Navigator for Cluster as a Storage
Node. The storage node provides all the Time Navigator agent functionalities as well as
media management capabilities.
Installation
The Time Navigator Storage Node for Cluster installation procedure is identical to the
Agent installation. See “To install Time Navigator for Cluster on the virtual machine”,
page 14 (Windows), or “To install Time Navigator for Cluster”, page 18 (Unix).
In the Installation Type screen, make sure that you choose Storage Node Installation.
Also make sure you are familiar with the Storage Node concepts and procedures described
in the Time Navigator in SAN Environment Chapter of the Time Navigator
Administration Guide.
42 Time Navigator for Cluster
43
Appendix
Appendix 1: FAQ
Q : How many machines must be created in Time Navigator
Administration Console to back up the cluster?
A : You must declare as many clients as there are virtual resources, i.e. cluster groups
containing disks to back up.
Principles
The procedure to restore a cluster is identical to restoring a standard Time Navigator agent.
You must first select the data to restore, synchronize in time and then start the restore.
However, if a switchover occurs while you are restoring, the restore process will stop on
error. You must wait until the switchover is complete and resume restoring.
Quorum Restore
Node Restore
To restore the physical nodes, you may use either a cold backup performed with
Time Navigator Disaster Recovery for Windows (Powered by WinPE), or a hot backup
performed with the System State application of the Time Navigator Agent for Windows
agent.
☞ See the Time Navigator Restore Guide and the Time Navigator Disaster Recovery for
Windows (Powered by WinPE) guide for details.
Cluster Restore
Automatically, Time Navigator is going to try to restore the Quorum on a shared resource
with the following name: \\Group_Virtual_Name\\CLUSTINA. However, at this point
in time, the cluster is not operational yet.
To reset the cluster in the Quorum, start the cluster service with the following option:
clussvc -debug noquorumlogging
Once started, the cluster uses the local node registry configuration.
You can then proceed with the Quorum restore using Time Navigator. The quorum is
restored in the TINA_HOME\Clustina directory.
SYMPTOMS
The Cluster service may not start if a hardware failure occurs or power is lost to both nodes
of a cluster and to the storage device known as the quorum on the shared device bus. In such
cases, the following error message may occur when you attempt to start the Cluster service
on the forming node of the server cluster:
CAUSE
This error message occurs if the quorum log file (Quolog.log) cannot be found or read at
Cluster service startup. It must be readable to confirm that the cluster configuration on the
local node is up-to-date. If it cannot read the log, the Cluster service does not start to prevent
potentially loading stale configuration data.
RESOLUTION
If you have a backup of the system state on one of the computers after the last changes were
Appendix 47
made to the cluster, you can restore the quorum by restoring this information. For additional
information about backing up and restoring cluster configuration information, see article
248998 "How to Properly Restore Cluster Information" in the Microsoft Knowledge Base.
If you do not have a backup of the Quorum log file, re-create a new quorum log file based
on the cluster configuration information in the local system's cluster hive by starting the
Cluster Service with the ResetQuorumLog switch. To do this, follow these steps:
1. Start the Services snap-in. (Click Start, point to Programs, click Administrative Tools,
and then click Services.)
2. Right-click and select the properties of the Cluster Service.
3. In the Start Parameters box, type:
-resetquorumlog
Then click the Start button.
If the error message occurs after you restore the system state on a computer that has lost the
quorum log, the quorum information is copied to %SystemRoot%\Cluster\Cluster_backup.
You can use the Clusrest.exe tool from the Resource Kit to restore this information to the
quorum disk. For more information about Clusrest.exe and other useful Resource kit
tools, see the Resource kit online Help.
MORE INFORMATION
Important: If possible, perform the procedure that is described in this article from the node
in the cluster that was known to have the latest configuration with regard to the resources
that are contained in this cluster. If you had a node that has not been a member in the cluster
since the time that the last changes, deletions, or additions were made to the cluster
resources, that node will not have the latest configuration. Using this node with the
ResetQuorumLog switch could result in permanent loss of data, such as Resources or
Resources Configurations.
The ResetQuorumLog switch is new for Windows 2000. This switch is not available in
Microsoft Windows NT 4.0. In Windows NT 4.0, if no quorum log is found,the Cluster
service assumes that it is a new installation and creates the quorum log fresh from the local
system information. It gets any updates cluster and program check points from the cluster
nodes when they are brought online.
48 Time Navigator for Cluster
Install Windows
1. Install exactly the same Windows version as the one backed up (same Windows
family, same language...).
2. Install all available Windows updates.
3. Include the node in the domain.
Note If you keep getting a blue screen, it could be an issue with the operating system,
either it is not the same as the one backed up or you did not install the latest
service pack and updates.
15. Since you restored the System State, you need to include the machine in the domain
again.
16. Verify that the cluster works correctly when a node switch occurs.
17. Set the environment variable TINA_ACD_FORCE_CLUSTER to No in the node System
State application.
50 Time Navigator for Cluster
Index 51
Index Node
installation 21
Failure 31
Q
Quorum
backup 11
restore 45
R
A Restore
Address, IP 7, 31, 43 method 35
Administrator, user 12 quorum 45
Alias 7 switching during a 35
B root
Backup user 18
method 33 S
quorum 11 Starting the daemon, Unix 19
switching during a 33 Storage disk arrays 7
C Switchover
Client during a backup 33
Managing clusters 31 during a restore 35, 45
Cluster failback 8
concept 7 failover 8
use 7 functional mode 33
D U
Disk arrays 7 Unix
E installation 18
Environment variables 15 starting the daemon 19
F User
administrator 12
Failback 8
root 18
Failover 8
I V
VMS
Installation
installation 20
NetWare 21
Unix 18 W
VMS 20 Windows
Windows 13 installation 13
IP
address 7, 31, 43
J
Job
backup method 33
restore mode 35
M
Managing clusters
Application modules 31
Client 31
N
NetWare
52 Time Navigator for Cluster