How can ADM Service customer, better monitor the agents from their end?

[Arvind Kandula]

How can ADM Service customer, better monitor the agents from their end?

[Arvind Kandula]

Best practice deployment and monitoring:

• To avoid SPOF (single point of failure), it is recommended to have a pair of agents per site and enable “Agent failover” (It is disabled by default).
• Ensure to map the 'Site Name' for each Agent. This will help ADM identify the agent accurately in the event of agent failover.
• Stagger ADM agent upgrades. Agents can be scheduled up to 24hrs apart. 
  • This can be scheduled in ADM under Infrastructure > Agents > Settings
  • From there you can specify upgrade times for all agents including local time zone
• Configure alerts for CPU, Memory, Disk & Agent Unreachable
  • From the same location you can configure notification settings for all of the above. E-mail, Slack, ServiceNow are all options. 
• Configure ADM to send out system level notification for Agent Upgrade & Agent Down
  • Setting > System Events > Notification  
  • Traps: AgentMonitoring & AgentDiag
• Optional: ADM can be configured for SNMP GET: https://docs.netscaler.com/en-us/citrix-application-delivery-management-software/current-release/manage-system-settings/create-snmp-manager-users.html 

Some other points to note:

• Always use the ADM UI to change password for ADM Agent 
• Resource requirement For ADM on-prem agent & ADM Service agent (recommended for better performance)
  • RAM: 32 GB, Virtual CPU:8 vCPU, Storage space: 30 GB
  • Acceptable latency between agent and ADC - 300ms
• For LSA agent - in case the LSA agent itself goes down, all the NetScaler instances connected to ADM Service for pooled licensing go into grace for one day. By the next day, ADM Service elects a new agent as LSA. Behavior is by default and does not impact functionality.