엮음 7.1
DB or unix user plugin.
zarafa-archiver and zarafa-archiver-extra package on RPM based distributions:
rpm -Uvh zarafa-archiver_<version>_<platform>.rpm zarafa-archiver-extra_<version>_<platform>.rpm
<version> with the correct version and <platform> with the required target platform (i386, i586, x86_64).
dpkg -i zarafa-archiver_<version>_<platform>.deb zarafa-archiver-extra_<version>_<platform>.deb
<version> with the correct version and <platform> with the required target platform (i386, x86_64).
/etc/zarafa/license directory of all your servers. Execute the following commands on every node to use the archive subscription:
echo 'Archiver code' > /etc/zarafa/license/archiverbase /etc/init.d/zarafa-licensed restart
/etc/zarafa/license directory, like normal ZCP CALs.
all server nodes, otherwise de-stubbing will not work.
/etc/zarafa/archiver.cfg.
zarafa-archiver needs to know where to connect to and how to authenticate. This is configured using the server_socket, sslkey_file and sslkey_pass settings.
server_socket = file:///var/run/zarafa sslkey_file = /etc/zarafa/ssl/client.pem sslkey_pass = secret
zarafa-archiver uses one central MySQL database for managing deduplication of archived attachments. The MySQL settings can be configured like this:
mysql_host = localhost mysql_port = 3306 mysql_user = zarafa mysql_password = password mysql_socket = mysql_database = zarafa-archiver
mysql> GRANT ALL PRIVILEGES ON zarafa-archiver.* TO 'zarafa'@'<ip-address_primary_node>' IDENTIFIED BY 'password'; mysql> FLUSH PRIVILEGES;
/etc/zarafa/archiver.cfg file. Three different modes for running the Archiver are available:
| Setting | Stubbing only | Hybrid | Archive only |
|---|---|---|---|
|
Writable archive
|
No
|
No
|
Yes
|
|
Archiving enabled
|
Yes
|
Yes
|
Yes
|
|
Stubbing enabled
|
Yes
|
Yes
|
No
|
|
Automatic deletes in primary store
|
-
|
Yes
|
Yes
|
|
Purge in archive store
|
-
|
Yes
|
Yes
|
stubbing only setup the following settings need to be configured in /etc/zarafa/archiver.cfg.
archive_enable = yes archive_after = 90 stub_enable = yes stub_after = 90 stub_unread = no delete_enable = no delete_after = 0 purge_enable = no purge_after = 0 cleanup_action = delete cleanup_follow_purge_after = no enable_auto_attach = yes auto_attach_writable = no
archive_enable = yes setting enables the archive operation, which is essentially the copying from the primary node to the archive node. The archive_after = 90 setting causes the archive operation to be performed 90 days after the message was delivered.
stub_enable = yes setting enables the stub operation, which is the operation on the message in the primary store where the body and the attachments are removed and a reference to the archive store is created. The stub_after = 90 setting causes the message to be stubbed immediately after the archive operation. Unread emails will not be stubbed automatically even when they are older than the configured lifetime.
delete_enable and purge_enable will be disabled as a user will not access his archive mailbox directly in this mode.
hybrid setup the following settings need to be configured in /etc/zarafa/archiver.cfg.
archive_enable = yes archive_after = 90 stub_enable = yes stub_after = 90 stub_unread = no delete_enable = yes delete_after = 365 purge_enable = yes purge_after = 3650 cleanup_action = store cleanup_follow_purge_after = yes enable_auto_attach = yes auto_attach_writable = no
archive only setup the following settings need to be configured in /etc/zarafa/archiver.cfg.
archive_enable = yes archive_after = 365 stub_enable = no delete_enable = yes delete_after = 365 purge_enable = yes purge_after = 3650 cleanup_action = store cleanup_follow_purge_after = yes enable_auto_attach = yes auto_attach_writable = yes
zarafa-archiver can attach archive stores automatically based on user attributes stored in LDAP or in Active Directory. When using this way of attaching stores, zarafa-archiver will create genuine archive stores on the server and attaches the user stores to these archive stores based on the information found in LDAP or in Active Directory.
enable_auto_attach setting must preferably be set to yes in /etc/zarafa/archiver.cfg:
enable_auto_attach = yes
zarafa-archiver can be run periodically to perform the auto-attach operation:
zarafa-archiver --auto-attach
zarafaUserArchiveStores attribute has to be modified. This is a multi value attribute, which needs to be set to the server name or server names of the servers that contain an archive store for the user.
dn: uid=user,ou=users,dc=example,dc=com objectClass: inetOrgPerson objectClass: organizationalPerson objectClass: person objectClass: top objectClass: zarafa-user objectClass: posixAccount cn: User gidNumber: 0 homeDirectory: /bin/false sn: User uid: user uidNumber: 1000 givenName: User mail: user@server.com userPassword:: e1NTSEF9VzlXV0U3N1NEcW54UkJ3SFJkQUYvVkhrUj zarafaAccount: 1 zarafaUserServer: userServer zarafaUserArchiveServers: archiveServer
zarafa-archiver -A with enable_auto_attach = yes or zarafa-archiver --auto-attach.
Zarafa Features tab.

Archiver feature and click Properties. This will pop up the dialog in which the server names of the servers on which an archive store should exist for the selected user or users.

zarafa-archiver -A with enable_auto_attach = yes or zarafa-archiver --auto-attach.
zarafa-archiver -u <user name> --list
zarafa-archiver --list-archiveusers
> zarafa-admin --details user1
Username: user1
Fullname: User 1
Emailaddress: user1@cluster.sio2
Active: yes
Administrator: no
Address book: Visible
Auto-accept meeting req:no
Home server: cnode-1
Last logon: 12/09/2011 03:41:32 PM
Last logoff: 12/09/2011 03:41:32 PM
Mapped properties:
PR_GIVEN_NAME User
PR_SURNAME One
PR_EC_ENABLED_FEATURES pop3
PR_EC_DISABLED_FEATURES imap
PR_EC_ARCHIVE_SERVERS cnode-2
Attached archives: 1
Root Folder in Archive - User 1 [Read Only]
Quota overrides: no
Warning level: unlimited
Soft level: unlimited
Hard level: unlimited
Current store size: 14.86 MiB
Groups (1):
Everyone
Archive details on node 'cnode-2':
Current store size: 114.68 MiB> zarafa-admin --details user1 --type archive --node cnode-2 Current store size: 114.68 MiB
> zarafa-admin --details archive
Username: archive
Fullname: Archive Store
Emailaddress: archive@cluster.sio2
Active: no
Administrator: no
Address book: Hidden
Auto-accept meeting req:no
Home server: cnode-2
Last logon: 12/09/2011 03:41:32 PM
Last logoff: 12/09/2011 03:41:32 PM
Mapped properties:
PR_GIVEN_NAME Archive
PR_SURNAME Archive
PR_EC_ENABLED_FEATURES pop3
PR_EC_DISABLED_FEATURES imap
Quota overrides: no
Warning level: unlimited
Soft level: unlimited
Hard level: unlimited
Current store size: 114.68 MiB
Groups (1):
Everyonetype and node arguments:
> zarafa-admin --unhook-store user1 --type archive --node cnode-2 Store unhooked.
type and node arguments:
> zarafa-admin --list-orphans --node cnode-2
Stores without users:
Store guid Guessed username Last login Store size Store type
-----------------------------------------------------------------------------------------------
F1A6BFCD67604B0FB733F746F1D00A91 user1 <unknown> 0 archive
> zarafa-admin --hook-store F1A6BFCD67604B0FB733F746F1D00A91 -u user1 --type archive --node cnode-2
Store hooked.type and node arguments:
> zarafa-admin --unhook-store user1 --type archive --node cnode-2
Store unhooked.
> zarafa-admin --list-orphans --node cnode-2
Stores without users:
Store guid Guessed username Last login Store size Store type
-----------------------------------------------------------------------------------------------
F1A6BFCD67604B0FB733F746F1D00A91 user1 <unknown> 0 archive
> zarafa-admin --remove-store F1A6BFCD67604B0FB733F746F1D00A91 --node cnode-2
Store removed.zarafa-archiver -A
--local-only option to zarafa-archiver causes it to only archive the primary stores that live on the server to which zarafa-archiver is connected. This is the server on which zarafa-archiver is executed unless otherwise configured in the configuration file.
zarafa-archiver -A --local-only
zarafa-archiver -u <user name> -A
/etc/crontab.
0 1 * * * root [ -x /usr/bin/zarafa-archiver ] && /usr/bin/zarafa-archiver -A
zarafa-archiver -C
--local-only option to zarafa-archiver causes it to only cleanup the archives of users who have a store on the server to which zarafa-archiver is connected. This is the server on which zarafa-archiver is executed unless otherwise configured in the configuration file.
zarafa-archiver -C --local-only
zarafa-archiver -u <user name> -C
/etc/crontab.
0 3 * * 0 root [ -x /usr/bin/zarafa-archiver ] && /usr/bin/zarafa-archiver -C


za-aclsync Synchronizes archive ACL settings with those of the primary store.
za-aclsync can be used to propagate the ACL settings from the primary store to the archive stores. However, no user will ever get more rights on a store or folder than the owner of the archive has. So if the archive was attached without write permissions, no user will get write permissions on the archive stores.
za-aclsync will first determine the rights for the owner of the archive. Then it will get all the entries from the ACL of the current folder except those of the owner. Each right will be masked with the rights of the owner before being added to the ACL of the archive folder.
za-aclsync [options] [users] options: -h serverpath : Host to connect to. -s sslkey_file : SSL key file for authentication. -p sslkey_pass : Password for the SSL key file.
users is a space separated list of users for which to synchronize the ACL settings. If no user is specified all users will be processed.
za-aclset sets or updates the permissions on an archive store for the owner of that store.
za-aclset [OPTIONS] [users...]
options:
-h | --host Host to connect to. Default: file:///var/run/zarafa
Three formats are allowed for this option:
UNIX socket : file://<path to the UNIX socket>
HTTP : http://<host or IP>:<port>/zarafa
Secure HTTP : https://<host or IP>:<port>/zarafa
-s | --sslkey-file SSL key file for authentication.
-p | --sslkey-pass Password for the SSL key file.
-w | Grant write permissions on the archive.
--writable <y|n> Enable or disable write permissions.
--help Show this help message.users is a space separated list of users for which to synchronize the ACL settings. If no user is specified all users will be processed.
za-restore can be used to repopulate a primary store to a state where no archive is required to read any message. In a less cryptic way this means that all stubbed messages are destubbed and all messages that were deleted after archiving are restored.
za-restore over dragging the messages back from the archive in Outlook or Webaccess is that the restored messages are sanitized, allowing them to be properly re-archived later.
Usage: za-restore [OPTIONS] user
OPTIONS:
-h | --host : Host to connect to. Default: file:///var/run/zarafa
-s | --sslkey_file : SSL key file for authentication.
-p | --sslkey_pass : Password for the SSL key file.
-l | --log-file : Specify log file.
--detach : Detach the selected or all archive stores before
starting the restore procedure. This avoids the
Archiver from rearchiving restored messages.
--unhook : Unhook the selected or all archive stores once
the restore process has completed. This implies
--detach and only works on archive stores.
--remove : Remove the selected or all archive stores once
the restore process has completed. This implies
--unhook and only works on archive stores.
--select-source : Select the source archive(s) by providing a comma
separated list of archive indexes. The indexes specify
which archives to restore from. The --detach, --unhook
and --remove options only apply to the selected
archives.
The archive indexes can be obtained by listing the
attached archives for a user: zarafa-archiver -u
<user> -l.
-v | --verbose : Increase console loglevel. Can be specified multiple
times.
-q | --quiet : Decrease console loglevel. Can be specified multiple
times.
-N | --dry-run : Don't actually modify anything.
--help : Show this help message./tmp/john_doe_restore.log
> za-restore --unhook -s /etc/zarafa/ssl/archiver.pem -p password \ -l /tmp/john_doe_restore.log john_doe
file:///var/run/zarafa. The sslkey_file and sslkey_pass are specified in order to connect to the other nodes in the cluster.