Veritas Access Installation Guide

To migrate NFS shares

1. Save the settings of the existing shares and use he same settings to recreate the shares on the destination file system.

   For NFS, scale-out file system uses NFS Ganesha (GNFS). You have to switch to the Kernel NFS (KNFS) server for the destination CFS.

   Use the following command to find out share information like nfs options, export dir and hostname.

   nfs> share show

2. Unexport the share using the following command:

   nfs> share delete <export_dir>

3. Stop the NFS server using the following command:

   nfs> server stop

4. Recreate the NFS on the destination CFS. Start the NFS server in KNFS mode. Switch to KNFS if it is in GNFS.

   nfs> server status

   nfs> server switch

5. Start the server.

   nfs> server start

6. Add the share to the destination CFS using the same settings as in the source scale-out file system.

   nfs> share add <nfsoptions> <export_dir> [client]

   For example:

   nfs> share add rw,async /vx/scaleoutfs1

To migrate CIFS shares

1. Save the existing configuration.

   cifs> show
   Name 																		Value
   ---- 																		-----
   netbios name					 					vmclus1
   ntlm auth 													yes
   allow trusted domains 	no
   homedirfs
   aio size 														0
   idmap backend 									rid:10000-1000000
   workgroup 													WORKGROUP
   security 														user
   Domain
   Domain user
   Domain Controller
   Clustering Mode 							normal
   Data Migration 								no

2. Make a note of the share name.

   cifs> share show
   ShareName    File System  Share Options
   =========== ========== 			==============================================
   mycifsshare lfs2 									owner=root,group=root,fs_mode=1777,rw,full_acl

3. Make a note of the homedir, if any.

   cifs> homedir show

4. Get the list of local users.

   cifs> local user show
   List of Users
   -------------
   admin
   user1

5. Get the local group.

   cifs> local group show
   List of groups
   -------------
   nogroup
   selftest
   mygrp

6. Create a new CIFS share on the destination CFS. Check the status of the CIFS server.

   cifs> server status

7. Start the CIFS server.

   cifs> server start

8. Add the file system to CIFS by specifying the name of the CFS, share name and modes.

   cifs> share add file_system sharename [@virtual_ip] 
   [cifsoptions]

   For example:

   cifs> share add mycifs myshare ro

   cifs> share show

9. Allow the user to access the share.

   cifs> share allow sharename @group1 
   [,@group2,user1,user2,...]

   For example:

   cifs> share allow myshare user1

To migrate the existing buckets on the scale-out file system

1. Get the name of the file system on which the bucket was created.

   objectaccess> bucket show
   Bucket Name FileSystem 				Pool(s) Owner
   =========== ============== ======= =====
   cloud1 					S3fs1611925982 mypool 	root

2. Create the new CFS.
3. Copy the data of the old file system on which the bucket was created to the new file system using the migration_tool.py script.
4. When you migrate theS3 bucket from a scale-out file system to CFS, the mapping of existing bucket to the new file system is not possible as it already exists in the S3 database. The unconfig_s3bucket.py script removes the existing entry of the scale-out file system from the S3 database to allow the mapping of the bucket to the new CFS directory path.

   ./unconfig_s3bucket.py fs_name

   where fs_name is the name of the file system to be removed from the S3 bucket mapping.

5. After the entry is removed from the S3 database, use the objectaccess map command to map the new CFS directory.

   objectaccess> map fs_path user_name

6. Delete the old scale-out file system bucket, if required. This operation does not delete the source file system.

To chage the S3 configuration for future buckets if the fs_type was set to largefs

1. Check the current settings of th eS3 server.

   objectaccess> show

2. If the fs_type is set to largefs, set the fs_type to the desired layout.

   objectaccess> set fs_type

To recreate the bucket:

1. Set the default pool.

   objectaccess> set pools pool_name

2. Enable the server.

   ObjectAccess> server enable

3. Start the server.

   ObjectAccess> server start

4. Set the file system size, as required.

   ObjectAccess> set fs_size size

5. Set the file system type and layout for the CFS.

   ObjectAccess> set fs_type layout

6. Create keys for user authentication and save the access key and secret key.

   /opt/VRTSnas/scripts/utils/objectaccess/objectaccess_client.py 
   --create_key --server ADMIN_URL --username root --password 
   P@ssw0rd --insecure

   The ADMIN_URL is admin.<cluster_name>:port. The port is 8144. This url should point to the Access Appliance management console IP address.

7. Map the bucket to the existing file system.

   Objectaccess> map fs_path	user_name