Backing Up and Reinitializing Hosted Unified Communications Services Components
This chapter explains how to backup and reinitialize the Hosted Unified Communications Services (Hosted UCS) platform components. It also provides some recommendations for upgrading to a newer version. This chapter includes the following sections:
•
Backing Up Cisco Unified Communications Manager and Cisco PGW
•
Restoring Cisco Unified Communications Manager and Cisco PGW Configuration
•
Backing up and Restoring USM
•
Clearing a Cisco Unified Communications Manager Cluster
•
Initializing the Cisco PGW
•
Initializing USM
Backing Up Cisco Unified Communications Manager and Cisco PGW
This section outlines the process for storing a known, reliable configuration before a platform upgrade and includes the following topics:
•
Backing-up Cisco Unified Communications Manager
•
Restoring Cisco Unified Communications Manager and Cisco PGW Configuration
•
Restoring the Cisco PGW Configuration
•
Restoring the Cisco PGW to Clean Status
After backup, the stored configuration can be restored onto a Hosted UCS platform component, if required. For example, restoring the initial static configuration for the Cisco Unified CM or Cisco PGW eliminates the time-consuming reconfiguration process.
Backing-up Cisco Unified Communications Manager
To use Disaster Recovery System (DRS) for backing up and restoring configuration, complete the following steps.
Procedure
Step 1
Go to Disaster Recovery System, using the Cisco Unified CM OS admin credentials. (CCM7.x).
Step 2
Create a backup device on the Cisco Unified CM, to add a network based backup device with access credentials.
Step 3
Choose Manual Backup > Select the Device Name.
Step 4
Check the CCM check box.
Step 5
Click Start Backup.
After approximately five to ten minutes, you should see the TAR file in Backup Device folder.
Backing Up the Cisco PGW
Note
If your Cisco PGW 2200 Softswitch is a continuous service system, (active-standby) ensure that you perform backup procedures on both Cisco PGW 2200 Softswitches.
Note
You can run the various backup operations that are described in the following sections only when you are logged in to your system as mgcusr. You cannot perform any backup operation while you are logged in as root.
To perform a manual backup operation, enter the following UNIX command on the Cisco MGC:
mgcbackup -d path [-r retries -t retry_time]
Where:
•
path—The full path of the directory in which to store the backup file; for example, a directory on a remote server that you have mounted on your system, or the local tape drive.
Note
Cisco recommends that you do not store backup files on your local Cisco MGC host, because storage of backup files on the local host reduces the amount of disk space available to process call data and does not ensure that the data is safe in the event of a natural disaster.
•
retries—The number of times to check for an active provisioning session on the Cisco MGC before aborting the backup operation. The default value is 0 and the maximum value is 100.
Note
A backup operation cannot start while there is an active provisioning session on the Cisco MGC.
•
retry_time—The number of seconds to wait between checks for an active provisioning session on the Cisco MGC. The default value is 30 seconds and the maximum value is 3600 seconds.
For example, to perform a manual backup operation where the backup file is saved to a directory path called /dev/rmt/h0, with a maximum of three attempts, each 60 seconds apart, you would enter the following UNIX command:
mgcbackup -d /dev/rmt/h0 -r 3 -t 60
The backup file is stored in the specified directory path in the following format:
mgc_hostname_yyyymmdd_hhmmss_backup
Where:
•
hostname—The name of the Cisco MGC host, such as MGC-01.
•
yyyymmdd—The date the backup file is created, in a year-month-day format, such as 20011130.
•
hhmmss—The time the backup file is created, in an hour-minute-second format, such as 115923.
For more information on backup operations, see the "Backing Up System Software" in Chapter 3 of the Cisco PGW 2200 Softswitch Release 9.8 Operations, Maintenance, and Troubleshooting Guide at the following URL:
http://www.ciscosystems.cg/en/US/docs/voice_ip_comm/pgw/9.8/Maintenance/Guide/r9chap3.html#wp1229244
Restoring Cisco Unified Communications Manager and Cisco PGW Configuration
This section explains how to restore the configuration for Cisco Unified CM and the Cisco PGW. It includes the following topics:
•
Restoring Cisco Unified Communications Manager Configuration
•
Restoring the Cisco PGW Configuration
•
Listing the Cisco PGW Backup Files
•
Restoring the Cisco PGW Backup File
•
Restoring the Cisco PGW to Clean Status
Restoring Cisco Unified Communications Manager Configuration
To restore Cisco Unified CM, follow the Disaster Recovery System (DRS) restore process and then restart the Publisher and the Subscribers.
Restoring the Cisco PGW Configuration
This restoration method uses a script to restore the configuration data for the Cisco MGC software, UNIX administrative files, and the Main Memory Database (MMDB).
Note
These procedures assume that you have backed up your system configuration data regularly. The procedures for system configuration backup can be found in Backing Up the Cisco PGW.
Listing the Cisco PGW Backup Files
To list the backup files in a particular directory path, enter the following UNIX command on the Cisco MGC:
Where path is the directory path in which you have stored backup files, such as a directory on a remote server or a local tape drive.
The system returns a response similar to the following:
Backup files in /var/cisco
--------------------------------------------------
mgc_venus_20011010_153003_backup
mgc_venus_20011011_153003_backup
mgc_venus_20011012_153003_backup
Restoring the Cisco PGW Backup File
CiscoMGC service on the PGW must be stopped before restoring a backup file.
Log on PGW as root user and run the following command to stop mgc service.
PGW-ENT2M% /etc/init.d/CiscoMGC stop
To restore the configuration data stored in a particular backup file, enter the following UNIX command on the affected Cisco MGC to run the restore script:
mgcrestore -d path -f filename
Where:
•
path—The directory path to the location where your backup files are stored.
•
filename—The file name of the backup file you want to restore.
For example, to restore a backup file called mgc_venus_20011012_153003_backup stored in a directory path called /var/cisco, you would enter the following command:
mgcrestore -d /var/cisco -f mgc_venus_20011012_153003_backup
After restoring the backup file, start the mgc service on PGW as root user.
PGW-ENT2M% /etc/init.d/CiscoMGC start
For more information on backup operations, see "Restoring Procedures for Cisco MGC Software Release 9.1(5) and up" in Chapter 5 of the Cisco PGW 2200 Softswitch Release 9.8 Operations, Maintenance, and Troubleshooting Guide at the following URL: http://www.ciscosystems.cg/en/US/docs/voice_ip_comm/pgw/9.8/Maintenance/Guide/r9chap6.html#wp1722661
Restoring the Cisco PGW to Clean Status
To clear the Cisco PGW and restore it to its original static configuration, complete the following steps.
Procedure
Step 1
Logon to the Cisco PGW using Reflexion Host - Unix.
On test systems: username=mgcusr. password=cisco.
Step 2
Enter the text shown in boldface:
GL-D-PGW mml> prov-sta::srcver="pure-static",dstver="iBVSconfig",confirm
MGC-01 - Media Gateway Controller 2005-06-30 14:09:55.352 BST
Note
pure-static in the MML command above is an example name of the static configuration MML session on the PGW. If you have static MML configuration session saved on your PGW, use that session name to restore PGW static configurations
MGC-01 - Media Gateway Controller 2005-06-30 14:10:02.164 BST
M COMPLD
"PROV-CPY"
;
In this example, the entry iBVSconfig is a temporary name. The exact name is not important.
Backing up and Restoring USM
The USM automatically backups the database within the cluster and between active and standby USM servers. USM always maintains four copies of the database, two in each headend. If a copy of the database needs to be saved offsite, you can set up an export copy of the database.
In most cases to date, offsite backup has occurred once every 24 hours. This should occur at a time of low provisioning traffic, such as in the early hours of the morning.
For the backup to be useful as part of a disaster recovery plan, the USM backup needs to be in a consistent state with those taken for the Cisco PGW and Cisco Unified CM, along with IP Unity if they are included in the Hosted UCS platform. To ensure a consistent state, there should be a USM transaction freeze while the platform is being backed up.
If all the backups are taken at the same time, it becomes possible to time-shift the entire platform back to the latest backup without any misalignment between USM and the servers that it controls.
Note
To use the CLI, a client capable of SSH is required. The most common SSH clients include Putty (Windows platform) or the SSH client in a terminal (UNIX / Linux or Mac OS X platform).
Note
The default login username is usmcli and the default password is voss. If you are unable to login using the default login details, please contact your dedicated support person so that they can reset the default password and verify that this functionality has been enabled for your system.
It is important to ensure that the commands below are carried out on the active VoSS server. These commands do not, and should not, be run on the inactive VoSS director.
To use the CLI, do the following:
Step 1
SSH to the cluster/standalone system and login as the user usmcli. You will be presented with a prompt as follows:
Welcome to the VOSS-USM Standalone Console
Step 2
To enter enable mode (similar to other networking hardware) type enable as follows:
The prompt should changes from "=>>" to "=>#" to reflect this change.
Step 3
Type "dbbackup" command on the CLI as follows:
Note
Only the database is backed up. Any other configuration information currently such as DNS Configuration, syslogs, Any Manual Changes (i.e. changes to file content), bulk loader sheets etc., must be backed up manually by copying those files.
Note
If you are rebuilding a system however, these files would be recreated during the installation of the platform and software.
The USM database backup is performed as shown in Figure 5-1.
Figure 5-1 Backup USM database using CLI
Restoring USM
Before restoring USM batabase backup, you must stop the relevant services on the USM system.
Procedure:
Step 1
Log on USM as root user.
Step 2
Stop all services and restart Postgress: (Standalone, on cluster environment you just need to restart postgress). In a Standalong USM system, run the following command to stop all the services:
for svc in usm-autoregister loader_scheduler extdhcp apache2 usm_batch selfcare memcached
estraier iptparent iptqueue iptdevmn voss-monit voss-imq postgresql-standalone postgresql;
do /etc/init.d/${svc} stop; done
If USM is deployed as a clutster, use the following command to stop the services:
/etc/init.d/heartbeat-wrapper stop
Restart postgresql services:
/etc/init.d/postgresql-stanalone restart
/etc/init.d/postgresql restart
Step 3
Perform the restore, see example below:
/opt/VOSS/usm/scripts/deployment/db-backup-restore.py --restore
/root/db-backup-ipt-2010-02-11.sql
Step 4
After restoring backup file start all USM services. In a Standalone USM system, run the following command to start all the services:
for svc in voss-monit iptdevmn iptqueue iptparent estraier memcached usm_batch selfcare
apache2 extdhcp loader_scheduler usm-autoregister; do /etc/init.d/${svc} start; done
If USM is deployed as a clutster, use the following command to start the services:
/etc/init.d/heartbeat-wrapper start
Clearing a Cisco Unified Communications Manager Cluster
This section describes the process for clearing a Cisco Unified CM cluster, in preparation to re-build the Hosted UCS platform.
The order of the clearing steps is not important and further clearing steps may be required on some Hosted UCS platforms. For example, you may need to delete organizations within the Movius server, using the Movius server Sysconfig GUI.
When you start the rebuild process, you must complete all stages. It is not possible to go back after you have cleared one component in the architecture.
You must clear the Cisco Unified CM before a rebuild to ensure that there will be no data duplication or mismatch between USM and the Cisco Unified CM.
You can quickly restore the Cisco Unified CM publisher to its initial state by restoring a CUCM backup file.
When no DRS backup file is available, to clear the Cisco Unified CM and avoid any interdependency issues, complete the following steps.
Procedure
Step 1
Delete phone devices.
Choose 50 phones for the search list to allow deletion of 50 phones at a time.
Step 2
From the Call Routing > Translation Patterns menu, delete Translation Patterns, 50 at a time.
Step 3
From the Device > CTI Route Points menu, delete any CTI route points used.
Step 4
From the Call Routing > Route/Hunt > Route Pattern menu, delete all route patterns.
Step 5
From the Call Routing > Route/Hunt > Route List menu, delete all route lists.
Step 6
From the RoutePlan > Route/Hunt > Route Group menu, delete all route groups.
Step 7
From the Device > Trunks menu, delete all trunks.
Step 8
From the Device > Gatekeepers menu, delete all gatekeepers.
Step 9
From the Device > Gateways menu, delete all gateways.
Step 10
From the Media Resources > MediaResourceGroupList menu, delete all media resource group lists.
Step 11
From the Media Resources > MediaResourceGroup menu, delete all media resource groups.
Step 12
From the System > Locations menu, delete all locations (show 50 at a time).
Note
The Cisco Unified CM does not allow to delete the default locations, device pools and regions. Do not attempt to delete default configurations.
Step 13
From the Media Resource > Conference Bridge menu, delete any conference bridges that are not required.
Keep the conference bridges that are required by USM.
Step 14
From the Application > CiscoCM Attendant Console > Pilot Points menu, delete pilot points used if any.
Step 15
From the System > Device Pool menu, delete all device pools, except Default.
Step 16
From the System > Region menu, delete all regions except Default.
Step 17
From the Call Routing > Route/Hunt > Hunt Pilot menu, delete all hunt pilots used.
Step 18
From the Call Routing > Route/Hunt > Hunt List menu, delete any hunt lists used.
Step 19
From the Call Routing > Route/Hunt > Line Group menu, delete any line groups used.
Step 20
Delete all users, either one-by-one via the CCMAdmin group or in bulk using the BAT Tool facilities.
Step 21
From the Call Routing > Directory Numbers menu, delete any directory numbers used.
Step 22
From the Call Routing > Call Pickup Group, delete any call pickup numbers used.
Step 23
From the Call Routing > Call Park menu, delete an call park numbers used.
Step 24
From the System > Cisco Unified CM Groups menu, delete all Cisco Unified CM groups except Default.
Step 25
From the Device > Device Settings > Device Profile menu, delete all profiles, including "Logout" service.
Step 26
From the Call Routing > Route Plan Report menu, search for unassigned DNs and select Delete All Found Items at the bottom of the search page.
This allows deletion of 150 unassigned DNs at a time.
Step 27
For voice-mail profiles, voice-mail pilot numbers, and MWI numbers, unless these need to be maintained.
Step 28
From the Call Routing > Class of Control > Calling Search Space menu, delete all CSSs except IncomingToCluster.
Step 29
From the Call Routing > Class of Control > Partitions menu, delete all partitions.
Step 30
From the Call Routing > Route Plan Report menu, search for assigned DNs and delete DNs one at a time.
Note
If issues occur, use the dependency record feature to search for components that might be preventing deletion of records.
Initializing the Cisco PGW
This section describes the clearing process for the Cisco PGW by deleting the USM-created file before rebuilding a Hosted UCS platform.
You must clear the Cisco PGW before reloading a Hosted UCS platform. Clearing the Cisco PGW means clearing out USM data but not other configuration information that may have been set up on the Cisco PGW servers independently of USM.
To initialize the Cisco PGW, complete the following steps.
Procedure
Step 1
Logon to the active Cisco PGW.
Log in over Telnet or SSH, using a terminal console program, such as PuTTY.
On test systems, the user account/password is mgcusr/cisco.
To configure the up arrow operate to add back previous lines, use the following:
Step 2
To verify that you are logged into the active Cisco PGW, enter the following commands:
Step 3
To create a binary backup to allow rollback if required, choose your own filename.
For example, 270710-01bin.
mml> prov-sta::srcver="active",dstver="270710-01bin"
Step 4
To create a text backup for diagnostics if required, choose your own filename.
For example,270710-01text.
mml>prov-exp:all:dirname="270710-01text"
Step 5
Restore the process, if rollback is required:
mml> prov-sta::srcver="270710-01bin ",dstver="270710-03bin "
mml> prov-dply (Dual server PGW platform)
or
mml> prov-cpy (Single server PGW platform)
Step 6
For the Cisco PGW reset process (dial plans only), enter the following commands:
% cd /opt/CiscoMGC/etc/cust_specific
This gets a list of files stored including Text file.
% cd /opt/CiscoMGC/etc/cust_specific/270710-01text
Step 7
Make a note for all four-character mml files loaded by USM.
For example, copy ICCM.mml, P974.mml, XXXX.mml, XXXX.mml into Notepad.
mml> prov-sta::srcver="active",dstver="270710-02bin"
mml> numan-dlt:dialplan:custgrpid="XXXX"
where XXXX is the name of each four-character mml file.
Step 8
Repeat this process until all XXXX.mml files have been deleted. If you hit a dependency, go to the next file and cycle through until all files are deleted.
Note
All the dial plans are not deletable manually due to interdependencies between dial plan. Cisco recommends to back up PGW static configuration and restore the PGW static configuration when PGW needs to be cleaned up.
Step 9
Reload the ICCM dial plan as an empty file:
mml> numan-add:dialplan:custgrpid="ICCM",overdec="YES"
mml> prov-dply (Dual server PGW platform)
or
mml> prov-cpy (Single server PGW platform)
Step 10
On completion, take a further backup of the Cisco PGW.
This will be the static configuration of the Cisco PGW if, for example, the Cisco PGW needs to cleared by deleting static settings.
Step 11
To create a binary back-up to allow rollback if required, choose your own filename.
For example, VSR2-151007Static-HB-01bin.
mml> prov-sta::srcver="active",dstver="2710001-01bin"
Initializing USM
This section explains how to clear an existing USM platform that has already been loaded with dial plans and data.
Clear the USM database when you are planning to rebuild the Hosted UCS platform. The clearing process is much faster than deleting all the data manually through the USM GUI and even faster than the Delete Bulk Loader tool or Operations tools. This is especially the case if USM has many customers and locations already loaded.
To clear a USM cluster, complete the following steps.
Caution
There is no way to undo a destroy procedure, so check whether you are on the correct server.
Note
The consequences of running this command are great. Please ensure that you fully understand the impact of running this command prior to using it
Procedure
Step 1
SSH to the cluster/standalone system and login as the user usmcli.
You will be presented with a prompt as follows:
Welcome to the VOSS-USM Standalone Console
Step 2
To enter enable mode (similar to other networking hardware) type enable as follows:
The prompt should changes from "=>>" to "=>#" to reflect this change.
Step 3
Type "reset app" command on the CLI as follows:
Really reset the application? ? [y/N] : y
Application reset complete.
Step 4
Reboot the USM system.
Note
You only need to destroy Standalone VOSS system or VOSSDir1 because USM automatically replicates to the other servers.