Cisco Cisco ACE Application Control Engine Module Weißbuch

Page 1 of 36

Guide

Cisco Application Control Engine Monitoring Guide

What You Will Learn

This document describes many of the ways in which you can monitor the Cisco

Application Control Engine (ACE)

and how to observe its performance and identify problem areas both before and after trouble occurs.

Overview

Cisco ACE is based on a purpose-built multinetwork processor architecture. All data received by the interconnect

at the switching fabric goes through a field programmable field array (FPGA), also known as the classification and

distribution engine (CDE). Although the Cisco ACE Module has a 16-Gbps throughput channel, the Cisco ACE

4710 appliance is bound by four 1 Gigabit Ethernet ports. The Cisco ACE 4710 uses one network processor. The

Cisco ACE30 Module is equipped with four network processors: two per daughter card (Figure 1). The CDE uses a

hashing algorithm to help ensure that a single connection reaches the same network processor for its duration.

Since the Cisco ACE 4710 has just one network processor, it does not use a CDE (Figure 2). Cisco ACE Modules

use Cavium Octeon (CN5860) processors. Each processor has 16 cores and runs at 600 MHz.

The network processors make up the data plane and are responsible for handling all the connection processing in

the Cisco ACE. The control plane is maintained separately with a dedicated SiByte processor. The control plane

manages the Cisco ACE configuration, health monitoring (probes), interactive management sessions (Telnet, SSH,

Simple Network Management Protocol [SNMP], and syslog), and other control functions and control traffic. For

additional information about Cisco ACE architecture and connection processing, please ask your Cisco Sales
representative for a copy of “Connection Handling Within the Cisco Ace Module Hardware Architecture.”