Key Features of Enterprise Server:
Shared Workspace
A shared workspace contains the data model used across all NetBrain systems – both in the client workbench and server processes. Data is stored in the Enterprise Server and replicated in each workbench. The shared workspace contains three types of data:
- Network topology and design related data
Such as L3/L2 topology, routing design, traffic filtering, etc. This data is periodically re-calculated based on raw data collected from the live network.
- Historical data
Such as historical routing tables, configuration files, etc.
- Administrative data
Such as credentials to access live networks
A NetBrain administrator will build the initial instance of a shared workspace from an OE Workstation. The data will then be uploaded to the Enterprise Server and updated automatically by server processes from that point on. A shared workspace is the shadow image of a live network.
Configuration and Routing Management
NetBrain Enterprise Server has built-in configuration and routing management features, including:
- Automated collection of configuration files and routing table
- Historical analysis of configuration files and routing table
- Template-based auto-configuration
If a user already has a configuration management solution such as Opsware or Ciscoworks in place, NetBrain can work with such systems through its northbound API. However, if users need to have historical routing analysis capabilities, using built-in features is recommended.
Recurring Network Benchmark
The benchmark process periodically collects data from the live network to update the shared workspace. This process is an extension of the configuration management process in that it collects additional data to create a snapshot of an entire network. Users can customize the time and frequency of the benchmark process.
For a large network, more network servers should be deployed to ensure the benchmark process is completed in time.
Network Re-Discovery
The network re-discovery process ensures that changes in the live network are reflected in the shared workspace. It is a process defined by an administrator and executed from the network server. All new devices discovered will be added to the system immediately, and un-responsive devices will be grouped together so they can be disabled by an administrator with a few clicks.
Network Proxy
Network Proxy service provides a probing point to the live network for each OE Workstation. Enterprise Server can provide proxy for:
- SNMP access
- Traceroute
- Ping
- Telnet/SSH (non-interactive mode)
For large networks where there are a number of pockets, like DMZ and isolated network segments, deploying a few network servers will greatly enhance the troubleshooting capability for OE Workstation.
User Management
Each NetBrain user is associated with a role, where proper privileges are defined. There are many levels of privilege available such as:
- Virtual network access inside NetBrain
- Live network access with non-interactive mode
- CLI access of live devices
- Discover and update access for the shared workspace
With this feature, enterprises may have different roles defined for design engineers, NOC engineers and consultants, conforming to the company’s security mandate.
3rd Party Integration
NetBrain’s northbound API and southbound API work with a number of existing network management systems, including:
- IBM NetCool ticket system
- HP Opsware configuration management system
- CiscoWorks RME
- Lucent Vitalnet performance management system
The benefit of 3rd party integration is the protection of an existing investment while leveraging NetBrain’s map-based automation capability. For example, in a fully integrated environment, any NetCool alarm can drive the automated creation of a NetBrain Q-map with one click. From the Q-map, historical performance data of vitalNet can be accessed interactively. In the backend, Opsware or CiscoWorks’ configuration files can dynamically feed into NetBrain, without repetitive polling.
Enterprise Scalability
NetBrain is scaled to support extremely large networks with either server federation or workspace partitioning.
Server federation divides the key backend process across a series of physical servers, which can then be deployed closed to the live network or users. Server federation can scale up to 50,000 network devices and millions of end systems per federated system.
Workspace partitioning divides a giant network into a few independent networks, each supported by a shared workspace. Workspace partitioning can both improve performance and manageability for large networks. In some organizations, this is also a security requirement. For managed service providers who are managing multiple networks at the same time, this approach can evolve into a multi-tenant deployment.
|
NETBRAINTECH.COM