Capacity and performance monitoring • Benefits of capacity and performance monitoring, page 1 • Busy hour definition, page 1 Benefits of capacity and performance monitoring Capacity and performance monitoring ensures that the current system is running within safe engineering limits within the customer's network. Understanding the current capacity of the system and monitoring this capacity profile on a regular basis allows both the customer and Cisco to understand the subscriber growth and behavior for future trending and capacity planning activities. Regular capacity and performance monitoring also ensures the capacity health and stability of the nodes, which helps to avoid capacity and performance exposures from occurring. Busy hour definition The first step of capacity planning and monitoring is to identify a Cisco HCS system busy hour. The busy hour is defined as the peak one-hour period during the day when the largest volume of subscriber traffic (in terms of CPU, memory utilization, or IOPS) is handled by a platform. You should collect at least one week of field data to determine a trend in subscriber behavior. To assist in the analysis, remove odd days in the analysis, such as weekends if the utilization is too low or days when utilization is atypically too high such as a holidays. To determine the busy hour • UC Applications— Look for the peak CPU utilization and peak memory utilization within a 24-hour period. Correlate the CPU and memory utilization busy hour with the traffic profile (most importantly the call attempts). Create a graph to visualize the trend and mark the CPU and memory utilization on the chart. • SAN—Look for the peak IOPS within a 24-hour period based on a delay criteria. Correlate the CPU utilization busy hour with the traffic profile (most importantly the call attempts). Create a graph and set the IOPS utilization on the chart. This point determines the node busy hour for IOPS. In the case of the IOPS, maintenance, backup, and upgrade activities should be considered in the busy hour determination. (In lightly loaded data centers the maintenance period might show a perceived busy hour due to the amount of IOPS required.) Capacity Planning Guide for Cisco Hosted Collaboration Solution 9.2(1)SU1 OL-30560-01 1 Capacity and performance monitoring Busy hour definition The following table provides lists the sources of system data that are available. Note DCNM is an optional component in Cisco HCS 9.2(1). If you are not using DCNM in your deployment, skip the tasks and sections related to DCNM. Table 1: System data availability Assurance management product Type of monitoring provided Monitored HCS devices Cisco Prime Unified Operations Manager Device Fault, Availability, Performance, and Diagnostics UC Applications and Network Devices (Cisco Unified Communications Manager, Cisco Unity Connection, IM and Presence Service, IP Phones, Cisco Unified Personal Communicator, CUCI MOC, SRST, ASR1000, ISR, Catalyst 3750) VMware vCenter Device Fault, Availability, and Performance ESXi Hypervisor, VMs, Nexus 1000v Unified Communications Domain Provisioning Manager Unified Communications Domain Manager, UC Applications Cisco HCM-F Provisioning Cisco HCM-F Service Inventory Provisioning UC Applications UCS Manager Device Fault, Availability, and Performance UCS Hardware, UCS6140 Hardware DCNM-LAN Device Fault, Availability, and Performance Nexus 7000, Nexus 1000v DCNM-SAN Device Fault, Availability, and Performance MDS 9000 The following tables provide a list of metrics that are important to monitor and trend on a regular basis. Table 2: System metrics System metrics Source Subscribers Cisco Prime Unified Operations Manager/Service Inventory Tenants/Customers Cisco Prime Unified Operations Manager/Service Inventory Capacity Planning Guide for Cisco Hosted Collaboration Solution 9.2(1)SU1 2 OL-30560-01 Capacity and performance monitoring Busy hour definition System metrics Source Busy Hour Call Attempts (BHCA) Cisco Prime Unified Operations Manager Locations Unified Communications Domain Manager Table 3: VMware metrics VM metrics Source CPU Utilization vCenter Memory Utilization vCenter IOPS vCenter CPU ready vCenter IOPS read instructions vCenter IPPS write instructions vCenter Table 4: Provisioning metrics Provisioning metrics Source Cisco Unified Communications Managers Service Inventory / HCM-F IM and Presence Service Inventory / HCM-F Cisco Unity Connection Service Inventory / HCM-F Cisco Prime Unified Operations Manager Service Inventory / HCM-F Cisco HCM-F Service Inventory / HCM-F Prime Central for HCS Service Inventory / HCM-F Unified Communications Domain Manager Service Inventory / HCM-F Cisco Unified Contact Center Service Inventory / HCM-F Table 5: Network infrastructure metrics Network infrastructure metrics Source VLANs CLI Capacity Planning Guide for Cisco Hosted Collaboration Solution 9.2(1)SU1 OL-30560-01 3 Capacity and performance monitoring Busy hour definition Network infrastructure metrics Source VLAN Port Instances UCS Manager VRFs CLI UCS Chassis CLI BGP Peers CLI Static Routes CLI HSRP Instances CLI OSPF Adjacencies (if applicable) CLI VMs vCenter Server Blades/Hosts vCenter Table 6: Performance metrics Performance metrics Source Call Success Rate Cisco Prime Unified Operations Manager By monitoring the capacity metrics, the bottleneck of the system will become evident, and proper capacity planning can occur to ensure that the bottleneck does not inhibit system resources or end-user experience. Capacity Planning Guide for Cisco Hosted Collaboration Solution 9.2(1)SU1 4 OL-30560-01 Capacity and performance monitoring Busy hour definition The following figure provides an example of how collecting and monitoring capacity metrics can assist in future planning. Figure 1: Capacity planning and forecasting Capacity Planning Guide for Cisco Hosted Collaboration Solution 9.2(1)SU1 OL-30560-01 5 Capacity and performance monitoring Busy hour definition Capacity Planning Guide for Cisco Hosted Collaboration Solution 9.2(1)SU1 6 OL-30560-01
© Copyright 2026 Paperzz