OSSEC is composed of multiple pieces. It has a central manager for monitoring and receiving information from agents, syslog, databases, and from agentless devices.
The manager is the central piece of the OSSEC deployment. It stores the file integrity checking databases, the logs, events, and system auditing entries. All the rules, decoders, and major configuration options are stored centrally in the manager; making it easy to administer even a large number of agents.
The manager may be called the OSSEC server, or even just server in this documentation.
The agent is a small program, or collection of programs, installed on the systems to be monitored. The agent will collect information and forward it to the manager for analysis and correlation. Some information is collected in real time, others periodically. It has a very small memory and CPU footprint by default, not affecting the system’s usage.
Agent security: It runs with a low privilege user (generally created during the installation) and inside a chroot jail isolated from the system. Most of the agent configuration can be pushed from the manager.
OSSEC can only be installed as an agent on Microsoft Windows platforms. These systems will require an OSSEC server, running on Linux or another unix-like system.
For systems that an agent cannot be installed on, the agentless support may allow integrity checks to be performed. Agentless scans can be used to monitor firewalls, routers, and even Unix systems.
OSSEC allows you to install the agent on the guest operating systems. It may also be installed inside some versions of VMWare ESX, but this may cause support issues. With the agent installed inside VMware ESX you can get alerts about when a VM guest is being installed, removed, started, etc. It also monitors logins, logouts and errors inside the ESX server. In addition to that, OSSEC performs the Center for Internet Security (CIS) checks for VMware, alerting if there is any insecure configuration option enabled or any other issue.
OSSEC can receive and analyze syslog events from a large variety of firewalls, switches and routers. It supports all Cisco routers, Cisco PIX, Cisco FWSM, Cisco ASA, Juniper Routers, Netscreen firewall, Checkpoint and many others.
This diagram shows the central manager receiving events from the agents and system logs from remote devices. When something is detected, active responses can be executed and the admin is notified.