Solution Prerequisites

The following are the solution setup prerequisites.

Hardware requirements

For HA deployment, each machine in the cluster should have the following hardware specifications.

Software Requirements

Installation Steps

Internet should be available on the machine where the application is being installed and connections on port 9242 should be allowed in the network firewall to carry out the installation steps. All the commands start with a # indicating that root user privileges are required to execute these commands. Trailing # is not a part of the command.

Allow ports in the firewall

For internal communication of docker swarm, you'll need to allow the communication (both inbound and outbound) on the ports: 7575/tcp,7676/tcp , 8090/tcp, 8091/tcp and 7077/tcp.

To start the firewall on CentOS (if it isn't started already), execute the following commands. You'll have to execute these commands on all the cluster machines.: 

# systemctl enable firewalld
# systemctl start firewalld

To allow the ports on CentOS firewall, you can execute the following commands. You'll have to execute these commands on all the cluster machines.

# firewall-cmd --add-port=7575/tcp --permanent
# firewall-cmd --add-port=7676/tcp --permanent
# firewall-cmd --add-port=8090/tcp --permanent
# firewall-cmd --add-port=8091/tcp --permanent
# firewall-cmd --add-port=7077/tcp --permanent
# firewall-cmd --reload

Configure Log Rotation

Add the following lines in /etc/docker/daemon.json file (create the file if not there already) and restart the docker daemon using systemctl restart docker. Perform this step on all the machines in the cluster in case of HA deployment.

{   
    "log-driver": "json-file", 
    "log-opts": {
        "max-size": "50m",
        "max-file": "3"
    } 
}

Creating Databases

Create two databases i:e (For UMM & ECM) in the MSSQL/MYSQL server with a suitable name and follow the application installation steps.

Installing Application

1. Download the deployment script ecm-deployment.sh and place it in the /root directory. This script will:
   1. delete the ecm-deployment directory in the present working directory if it exists.
   2. clone the ecm-deployment repository from gitlab in the present working directory.

2. To execute the script, give it the execute permissions and execute it. 

   # chmod +x ecm-deployment.sh
   # ./ecm-deployment.sh

3. Update environment variables  in the following files inside /root/ecm-deployment/docker/environment_variables folder.

   1. environment-variables.env

      Name	Description
      Do not change the default values for non-HA deployment. For HA, use SQL server cluster settings instead of the defaults.
      DB_URL	Database connection url For example: jdbc:jtds:sqlserver://<MACHINE-IP or FQDN>:PORT/db_name jdbc:jtds:sqlserver://<MACHINE-IP or FQDN>:PORT/db_name;instanceName=SomeInstance
      DB_USER	database user
      DB_PASS	database password
      DB_DRIVER	JDBC driver e.g., net.sourceforge.jtds.jdbc.Driver
      DB_DIALECT	Database dialect e.g., org.hibernate.dialect.SQLServer2008Dialect
      Change the following variables for ECM services
      DATABASE_ENGINE	sqlServer
      ECM_MODE	CCX This is cisco contact center type either UCCE or UCCX
      SYNC_ENABLED	YES
      SYNC_SCHEDULE	0 0/3 * 1/1 * ? * This is cron expression for sync process scheduler.
      FEED_SCHEDULE	0 0/1 * 1/1 * ? * This is cron expression for feed process scheduler.
      CISCO_DB_URL	UCCE awdb url, in case of UCCX it is UCCX database url For example: jdbc:jtds:sqlserver://<MACHINE-IP or FQDN>:PORT/db_name jdbc:jtds:sqlserver://<MACHINE-IP or FQDN>:PORT/db_name;instanceName=SomeInstance
      CISCO_DB_USER	Cisco database username
      CISCO_DB_PASS	Cisco database password
      MIN_IDLE	5, do not change.
      MAX_IDLE	20, do not change.
      MAX_ACTIVE	50, do not change.
      FETCH_DELAY	2, do not change.
      OUTDATED_INTERVAL	24, this is number of hours a contact is closed in ECM with Dangling status if it is not synced
      PCS_URL	http://<MACHINE-IP or FQDN>:PORT
      PCS_USERNAME	database username
      PCS_PASSWORD	database password
      REDUNDANT_DEPLOYMENT	Decides if synchronizer deployment is redundant. Set it to "true". Holds strings "true" or "false". Default is false
      INSTANCE_NAME	Used for differentiation of instances when deployed redundant, could be any string. It should be different on both machines.
      UCCE_SHARED_PATH_DOMAIN	Dialer import folder machine domain
      UCCE_SHARED_PATH_USER	Dialer import folder machine username
      UCCE_SHARED_PATH_PASSWORD	Dialer import folder machine password
      UCCE_SHARED_PATH_IP	Dialer import folder machine IP
      Change the following variables as per your environment For UMM
      PRIM_FINESSE_IP	Primary Finesse URL including port (if not 80 or 443) For example: http://finesse10-5.com https://finesse12-0.com:8445
      SEC_FINESSE_IP	Secondary Finesse URL including port (if not 80 or 443) For example: http://finesse10-5.com https://finesse12-0.com:8445
      FINESSE_USER	Finesse administrator user
      FINESSE_PASS	Finesse administrator password
      UMM_DB_URL	UMM Database
      UMM_DB_DRIVER	UMM Database driver
      UMM_DB_DIALECT	UMM Database dialect
      UMM_DB_PASS	UMM Database password
      UMM_DB_USER	UMM Database username
      ADMIN_PASS	The password of the admin user
      Change the following variables for Frontend
      GAT_URL	Provide the URL of Umm : http://<machine_IP>:<PORT>/<microservice> (microservice = umm) In the case of HA, machine_IP should be the virtual IP of HA
      CISCO_TYPE	UCCE/UCCX (capital case)

4. Get domain/CA signed SSL certificates for ecm FQDN/CN and place the files in /root/ecm-deployment/docker/certificates folder. The file names should be server.crt and server.key.

5. Copy the ecm-deployment directory to the second machine for HA. Execute below command 

   # scp -r ecm-deployment root@machine-ip:~/

6. Go to the second machine and update the environment variables where necessary like INSTANCE_NAME and GAT_URL

8. Execute the following commands inside /root/ecm-deployment directory on both machines.

   # chmod 755 install.sh
   # ./install.sh

9. Run the following command to ensure that all the components are up and running. The screenshot below shows a sample response for a standalone non-HA deployment. 

   # docker ps

   
   

Virtual IP configuration

Repeat the following steps for all the machines in the HA cluster.

1. Download keepalived.sh script and place it in /root directory.

2. Give execute permission and execute the script: 

   # chmod +x keepalived.sh # ./keepalived.sh

3. Configure keep.env file inside /root/keep-alived folder

   Name	Description
   KEEPALIVED_UNICAST_PEERS	IPs of the machines in the cluster. On each machine, this variable should have a list of IPs of all the other machines in the cluster. The format of the list is as below:  192.168.1.80
   KEEPALIVED_VIRTUAL_IPS	Virtual IP of the cluster. It should be available in the LAN. For example: 192.168.1.245
   KEEPALIVED_PRIORITY	Priority of the node. Instance with lower number will have a higher priority. It can take any value from 1-255.
   KEEPALIVED_INTERFACE	Name of the network interface with which your machine is connected to the network. On CentOS, ifconfig or ip addr sh will show all the network interfaces and assigned addresses.
   CLEARANCE_TIMEOUT	Corresponds to the initial startup time of the application in seconds which is being monitored by keepalived. A nominal value of 60-120 is good enough
   KEEPALIVED_ROUTER_ID	Do not change this value.
   SCRIPT_VAR	This script is continuously polled after 2 seconds. Keepalived relinquishes control if this shell script returns a non-zero response. It could be either umm or ECM backend API. pidof dockerd && wget -O index.html http://localhost:7575/

4. Give the execute permission and execute the script: 

   # chmod +x keep-command.sh # ./keep-command.sh

Adding License

1. Browse to http://<MACHINE_IP or FQDN>/umm in your browser (FQDN will be the domain name assigned to the IP/VIP). 
2. Click on the red warning icon on right, paste the license in the field and click save.