Archive:SEP sesam Extension für Scalix
SEPsesam for Scalix Groupware Server
(C)SEP AG, R. Seyerlein
Copyright 1999-2008 by SEP AG. All Rights reserved.
Any form of reproduction of the contents or parts of this manual is allowed only with the express written permission from SEP AG. When compiling and designing user documentation SEP AG uses great diligence and attempts to deliver accurate and correct information. However, SEP AG cannot issue a guarantee for the contents of this manual.
Introduction
SEP sesam online module for Scalix provides an online backup of Scalix Groupware Servers. This includes saving and restoring single user mailboxes; and full backup of the Scalix module, and in the event of a complete system failure, a complete Groupware database restore.
System requirements
- Scalix Server >=9.4 or greater on Linux
- Sesam Server >= 3.0.1.16 or greater
- Sesam DB-Online-Module for Scalix
- Previous installation of the standard Sesam Backup Client
- working SEP sesam file system backup for this client
- for complete online backups, Scalix data filesystem should be on an LVM volume providing snapshots.
Installation
- First, verify that the standard Sesam Backup Client is already installed
- File system backup must be active for this sesam client
- Login as root user
- Unpack the file <OS>_scalix_x.x.x.x.tgz in a temporary location
- Copy scalix/sbc_scx scalix/sbc_scxend into <SESAM_ROOT>/bin/sesam directory
root@scalix:~> cd /tmp root@scalix:/tmp> tar xvzf linux_scalix.tgz root@scalix:/tmp> cp sesam_scalix_3.0.1.27/sbc_scx /opt/sesam/bin/sesam root@scalix:/tmp> cp sesam_scalix_3.0.1.27/sbc_scxend /opt/sesam/bin/sesam
Backup Configuration
The further configuration can be accomplished using the SEP sesam GUI.
Configuration of a full Scalix backup
Full backup will be executed after reconciliation with Scalix. The Scalix server will be suspended to make a filesystem snapshot. After this the server operation is resumed. At this point Scalix data is then saved to the sesam server from the client.
- Create a new backup task for the Scalix server
- Choose Scalix as task type
- Enter path or browse via GUI
all:path=<path of Scalix data>:vgpath=<LVM device>:snapsize=<size of snapshot>
Example:
all:path=/var/opt/scalix:vgpath=/dev/scalixvg/scxlv:snapsize=200M
- Schedule the task like any other backup
- path
- Scalix data directory (mandatory)
- vgpath
- path name of the LVM volume (mandatory)
- snapsize
- size of the snapshot volume
- instance
- Scalix instance: only needed, if you are running more then one Scalix instance on the same server
The old backup source format all:<path of Scalix data>:<path of LVM device>:<size of snapshot> is still supported.
It's also possible to make an offline backup. In this case, sesam will shutdown Scalix
services completely and save the Scalix data directory directly without a file system snapshot. After backup the Scalix services are restarted. There is no access to Scalix services during this backup operation.
When using SuSE Linux Enterprise Server 10 you must enter the following command before backup. "/etc/init.d/boot.udev stop" You may also start this automatically before backup begins by entering the command in sbc_pre. (See SEP User Manual for Instructions) If you don't follow either of the steps above the creation of the Snapshot will fail. After the backup you must run the following command "/etc/init.d/boot.udev start". If you want to start this action automatically insert the command into sbc_post |
Configuration of mailbox backup
- Create a new backup task for the Scalix server
- Choose Scalix as task type
- Enter backup source in the following format:
mail:host=<mailserver>:instance=<Scalix instance>:mn=<mailnode>[:port=<LDAP port>] [,<user1>, <user2> ...]
- host
- hostname of mailserver (mandatory)
- instance
- Scalix instance: only needed, if you are running more then one Scalix instance on the same server
- mn
- name of the mail node (mandatory)
- user1 ...
- user name for mailboxes to saved. If no users are given all mailboxes of mailnode are saved
- port
- Listen port for Scalix LDAP servers. Only necessary if your not listening on default port 389
Replace the ',' in the name of mailnode by '.' This is necessary because Sesam uses ',' internally. The '.' will be automatically converted back to the Scalix notation during backup |
- If no users are given, all are saved.
- Schedule the task like any other backup (See SEP sesam User Manual for Details)
The format of backup source has changed from version 2.4. The old format
mail:<name of server>:<name of mailnode>, [<user1>, <user2>, ...]
still works, but the new form is preferred.
Example:
[root@ozeanix ~]# omshowmn ** ozeanix,sep
then the source for a backup of all user mailboxes and only on Scalix instance would be:
mail:host=ozeanix:mn=ozeanix.sep, sxadmin
If you have more Scalix instances:
mail:host=ozeanix:instance=ozeanix:mn=ozeanix.sep, sxadmin, Tux Pingi
For backup of mailboxes of users Tux Pingi, sep sesam and test user:
mail:host=ozeanix:instance=ozeanix:mn=ozeanix.sep, Tux Pingi, sep sesam, test user
Backup all mailboxes:
mail:host=ozeanix:instance=ozeanix:mn=ozeanix.sep
Exclude list
Exclude list can be used to omit certain mailboxes from backup. Exclude patterns are regular expressions which are compared against e-mail addresses.
Example:
support@scalix, admin@sep.de
will exclude
support@scalix.com support@scalix.sep.de admin@sep.de
Attention: Do not start both backup types at the same time, because they will block each other. |
Restore Configuration
Complete restore
A complete restore can only be done, if the Scalix services are offline. Because many files are interdependent, a full restore of the full saveset with option overwrite is recommended. Start the restore wizzard and perform the same steps as for a normal filesystem restore.
Bevore starting a full restore of the Scalix server, it is necessary to stop all Scalix processes!
Single mailbox restore
During restore of a single mailbox or multiple mailboxes, data will be restored to the Scalix server and then be imported. As target, a (restore) mailnode can be specfied into which data should be imported e.g. recover,scalix,com. This is useful to recover deleted e-mails without overwriting new ones. Because of license restrictions in Scalix 10, the restore mail node must be a sub domain of the original instance.
Sesam tries to create the mailnode. An already existing one will be reused. Then the users of the original mailbox will be added. If a user already exists and restore is started without option overwrite, mailbox will not be imported. The password of newly created users will be scalix.
Example for target path:
mail:host=ozeanix:instance=ozeanix:mn=restore.ozeanix.sep
Attention: If users are created in a different mailnode, you will not see them in the Scalix Admin Console. To delete them you have to use omdelu command. But access via webmail or IMAP is certainly possible.
Literature
For further reading on the topic we recommend the following documents:
- Scalix Backup & Recovery