How to migrate eDirectory from a NetWare 6.5 server to an OES2 Linux server using the migedir utility

The eDirectory migration process using migedir

• This tool does not migrate volumes or user's data from a NetWare server to an OES2 Linux server.
• IP address and DNS migrations are not performed by the eDirectory migration utility.
• Only the eDirectory instance will be migrated. Applications depending on eDirectory will not.
• Only the target server will be available after the migration. The source server will be locked.You should not use this migration methodology if both servers need to be available during the migration operation.

The eDirectory migration from NetWare requires both the migration of eDirectory data and the server identity to provide seamless accessibility after the migration. The eDirectory migration utility, provided with eDirectory 8.8 SP2 (the migedir command line utility), performs:

• Pre-migration tasks and health validations
  

1. The health and state of the replicas in the rings are verified.
2. Configuration information for the server being migrated is collected and written to a configuration file to be used by other operations during the migration.
3. Time synchronization is verified between the source and target servers.
4. The target server is checked for any existing eDirectory instances. If an instance exists, the user is prompted to allow the deletion of the instance. The existing instance is removed before proceeding with the migration.

• eDirectory backup on the source server and migration to the target server.
  • The utility migrates the eDirectory instance based on the collected configuration information. This involves backing up the source server's data, locking the eDirectory instance on the source server, migrating data (all eDirectory objects) to the target server, and restoring the eDirectory instance on the target server. Dependent NICI files are also migrated.
  • The utility also configures the local instance on the target server with the source server details obtained during the previous checks.

• Post-migration tasks After migration, the following tasks are performed by the utility:
  • The nds.conf configuration file is modified with the source server eDirectory instance information, such as tree name and server name.
  • The eDirectory instance on the target server is restarted so it can use the new data.
  • Network address repair is performed to start the synchronization of the new IP address in the replica ring.

Important: Even though eDirectory migration using migedir is tested and certified by Novell test labs, it is not a complete solution for migrating from NetWare to Linux. The migedir utility is intended as a standalone migration utility to be used when you want to migrate eDirectory alone and can't be used in concert with any of the other OES Migration Tools. Once the migedir utility is used, the services on the source server (NetWare 6.5) are no longer available. Compatibility issues with other utilities are scheduled to be addressed in OES 2 SP1, but until then you must choose between using migedir and using other OES Migration Tools.

• Source NetWare Server
  • The source NetWare server should be running and should not be part of any partition operation at the time of the migration.
  • The source has to be eDirectory 8.7.3 IR5 or greater
  • Copy the NFK file from the install NetWare 6.5 CD to sys:\system and rename it to sys:\system\nicifk
    

• Linux Target Server
  • The target server must be running OES 2 Linux.
  • Do not install any OES2 components on the target server except for eDirectory.
    
  • eDirectory 8.8 SP2 needs to be installed. You can install and configure eDirectory through YaST | Open Enterprise Server | OES Install and Configuration.
  • The default eDirectory 8.8 SP2 instance must already be configured and be active (this instance will be overwritten during the migration).
  • The target Linux server must be able to access the NetWare server remotely. (The eDirectory migration utility runs only on the target server.)

Valid Platforms:

The eDirectory migration utility is designed to run only on OES 2 Linux, which is the target platform for migration. Hardware and supported platform requirements are the same as those for OES 2 Linux.

1. Run the migedir utility by entering the following command on the target server:      migedir -s <IP address> [-A <log directory name>] [-t] [-v] [-h] The utility takes the following command line options:

   Option	Description
   -s <IP address>	Specifies the IP address of the source server containing the eDirectory instance to be migrated. This is a mandatory parameter.
   -A directory name	Enables auditing. Directory name specifies the directory in which log files should be created.
   -t	Tests the validity of the input parameters. This option verifies the IP address; however, it does not perform the actual migration.
   -v	Enables the verbose mode.
   -h	Prints help about using this utility.

2. Follow the on-screen display as the utility performs the migration.

Handling Failures

During migration, the database on the source server is locked to avoid running multiple copies of the instance on the source and target servers at the same time. Running multiple instances can lead to data inconsistency. If the process fails, and if you intend to bring up the source server again, you need to perform the following tasks:

• Remove the partially migrated eDirectory instance on the target server.Refer to "Removing a Server Object and Directory Services from a Tree" in the Novell eDirectory 8.8 Administration Guide for more information.

• Restore and unlock the database in the source server. The database backup is saved in the sys:ni/data folder.Refer to Section 15.0, "Backing Up and Restoring Novell eDirectory" in the Novell eDirectory 8.8 Administration Guide for more information.

Post Migration Procedures

After migration, the target eDirectory instance listens on the IP address of the target server and not on the source server’s address. Allow additional time after migration for the eDirectory instance to synchronize the new IP address in the replica ring. Successful eDirectory migration can be verified by performing eDirectory operations on the new IP address.

Important: If you want to use the existing security certificates, you must change the IP address of the target server to that of the source server. If you don’t want to do this, you must issue new certificates.

If you change the IP address of the target server after migration, you must modify the nds.conf file, restart the eDirectory instance, and repair the network address and partitions replica manually. For more information on repairing eDirectory instances, refer to Section 11.9, "Advanced DSRepair Options" in the Novell eDirectory 8.8 Administration Guide.

+ Advanced eDirectory Troubleshooting