SNMP Agent

• General
• Master-Subagent architecture

General

1. What are the SNMP versions supported by the SNMP agent?
2. Can multiple management stations (e.g., EMS + direct SNMP monitoring access) simultaneously access the SNMP agent?
3. Can I send SNMP v1 request to an SNMP v2 agent?
4. Can I prevent particular management station(s) from accessing my agent?
5. Can SNMP agents be remotely accessed by standard management consoles such as HP Openview, TNG Unicenter, etc.?
6. Is it possible to send different versions of trap to different management stations or NMS, from the SNMP agent?
7. What are the different trap indexes in SNMP trap/notification?
8. What is the difference between TRAP-TYPE and NOTIFICATION-TYPE in SNMP agent?
9. Does the SNMP agent toolkit provide database management through SNMP?
10. What does a node in the MIB Tree refer to?
11. Is it possible to port the agent developed by SNMP Agent Toolkit to devices with very limited resources?
12. What is Rule Engine?

Master-Subagent Architecture

1. How does WebNMS SNMP Agent Toolkit provide distributed management system in SNMP agent?
2. What is Master Agent and Subagent?
3. Can another vendor's SNMP agent be registered as subagent to WebNMS Master Agent?
4. Can WebNMS SNMP agent be made as subagent to native SNMP agents of Linux, Solaris, and Windows?
5. Should both the native agent and WebNMS SNMP agent be on the same box?
6. Does WebNMS support AgentX for master sub-agent architecture?
7. Can we run multiple sub-agents on a CPU?
8. Can a sub-agent be queried by multiple manager applications through master agent?
9. I don't want to expose the proxy configuration details of the sub-agents to remote management applications. Is it possible using SNMP Agent Toolkit?
10. Can I add sub-agent to Master Agent at runtime without bringing other agents down?
11. Does the toolkit support MIB II implementation?

General

1. What are the SNMP versions supported by the SNMP agent?

The SNMP agent supports SNMPv1, SNMPv2c, and SNMPv3 versions.

Questions

2. Can multiple management station (e.g., EMS + direct SNMP monitoring access) simultaneously access the SNMP agent?

Yes, any number of managers can access the SNMP agent. There is no restriction as to how many managers can access an SNMP agent. 

Questions

3.Can I send SNMP v1 request to an SNMP v2 agent?

Yes. SNMPv1/v2 request can be sent to an SNMPv2 agent

Questions

4. Can I prevent particular management station(s) from accessing my agent?

This can be achieved by using the Access Control Table associated with the agent, where authentication information are given. To restrict access to particular management stations refer to "Configuring Managers to Community" page of the help documentation.

Questions

5. Can SNMP agents be remotely accessed by standard management consoles such as HP Openview, TNG Unicenter, etc.?

Yes. SNMP agents can remotely be accessed by standard management consoles by just connecting to the port where the agent is running.

Questions

6. Is it possible to send different versions of trap to different management stations or NMS, from the SNMP agent?

SNMP agents can send different version traps to different destinations using Trap Forwarding Table. The table provides a column named ManagerVersion wherein the Trap version can be specified. The trap thus gets forwarded to the management stations according to the entries in the Trap Forwarding Table.

Questions

7. What are the different trap indexes in SNMP trap/notification?

It varies from 0-6, where 0-5 are generic traps.

coldStart (0), 
warmStart (1), 
linkDown (2), 
linkUp (3), 
authenticationFailure (4), 
egpNeighborLoss (5) and

Value 6 is for specific trap

Questions

8. What is the difference between TRAP-TYPE and NOTIFICATION-TYPE in SNMP agent? 

This table gives the difference between the SNMPv1Trap and SNMPv2c Notification.

Trap	Notification
SMI v1 macro	SMI v2 macro
Contains SNMP agent address.	Does not contain SNMP agent address.
The identity of of v1 trap is determined by examining the fields Enterprise,Generic Type and Specific Type.	The identity of v2c notification is determined by value of the snmpTrapOID(defined in RFC 3418) varbind.
TimeStamp field is present to denote time elasped since last re-initialization of agent.	sysUpTime(defined in RFC 3418) varbind denotes the time elasped since last re-initialization of the agent.
It does not have Request ID	It has Request ID
Does not support confirmed trap.	Supports confirmed notification (INFORM).

Questions

9. Does the SNMP agent toolkit provide database management through SNMP?

Yes. SNMP agent can access the database table and retrieve the stored information from it. This helps the user to save considerable memory space as database serves as a resource to persist data. For more details on database supported by the SNMP Agent Toolkit, refer to "Supported Storage Types for Scalars and Tables" of the help documentation.

Questions

10. What does a node in the MIB Tree refer to?

A node in the MIB Tree refers to an object to be managed by the agent. Each object whether it is a device or a characteristics of a device, must have a name by which it can uniquely be identified. The name is called the Object Identifier (OID).

The OID is written as a sequence of integers, separated by periods. For example, the sequence .1.3.6.1.2.1.1.1 specifies the system description within the system group of the management sub-tree.

Questions

11. Is it possible to port the agent developed by SNMP Agent Toolkit to devices with very limited resources?

Using WebNMS SNMP Agent Toolkit, it is possible to develop J2ME SNMP v1/v2c agents. They are compatible with Sun J2ME CDC/ IBM J2ME CDC specifications. The J2ME agents can be ported to devices with limited resources(like embedded devices having low memory footprint, etc). For more details, refer the "Developing and Porting SNMP Agent in J2ME CDC Environment" page in our help documentation.

Questions

12. What is Rule Engine?

WebNMS SNMP Agent Toolkit introduces the concept of Rule Engine using which users' business logic can be defined as a set of rules in configuration files and some tasks (like sending e-mails, traps, etc) can be performed based on this without modifying the code. Thus, the business logic need not be hard-coded in the stub files. For more details, refer the "Implementing Business Logic using Rule Engine" page in our help documentation.

Questions

Master-Subagent Architecture

1. How does WebNMS SNMP Agent Toolkit provide distributed management of SNMP agents?

WebNMS SNMP agent supports master-subagent architecture that provides distributed management capability.

Here, WebNMS master agent can act a proxy to other agents. Any agent that understands pure SNMP can be made as a sub-agent to the WebNMS's master agent.

You can go through the "Implementing SNMP Proxy" section in the product help documentation to know more about master-subagent concepts.

Questions

2. What is Master Agent and Sub Agent?

Sub agent:

The agents which are not directly exposed to the manager are called sub-agents.

Master agent:

The agent which acts as a proxy/interface for sub agent(s) is called master agent. It contains the details about sub agents and acts as a interface between the manager and sub agent. When SNMP requests intended for sub agents reaches the master agent, the master agent will forward those requests to the sub-agent and the responses from the sub agents will be sent back to the manager.

Questions

3. Can another vendor's SNMP agent be registered as sub-agent to the WebNMS master agent?

Yes. For further details, please refer to the Sub-agents section in the help documentation.

Questions

4. Can WebNMS SNMP Agent be made as sub-agent to native SNMP agents of Linux, Solaris, and Windows?

Yes. You can make Solaris, Linux, and Windows native SNMP agent as WebNMS agent's sub-agent. Refer Making WebNMS Agents as Subagents to Third-Party Agents section in our help documentation.

Questions

5. Should both the native agent and WebNMS SNMP agent be on the same box?

It is not mandatory to have both the native agent and the WebNMS SNMP agent in the same machine. This can be decided based on your requirements.

Questions

6. Does WebNMS support AgentX for master subagent architecture?

No, but WebNMS provides a better solution for master subagent architecture via SNMP i.e, the communication between master and sub-agent through SNMP enhances more flexibility as any legacy SNMP agent can be registered as sub-agent to WebNMS master agent and also WebNMS's SNMP agent can be registered as sub-agent to the existing master agent.

Questions

7. Can we run multiple SNMP sub-agents on a CPU?

Yes, you can start many SNMP sub-agents in the same CPU at different port numbers.

Questions

8. Can a sub-agent be queried by multiple manager applications through master agent?

Yes.

Questions

9. I don't want to expose the proxy configuration details of the sub-agents to remote management applications. Is it possible using SNMP Agent Toolkit?

It is possible to restrict the remote access of sub-agent configurations in the master agent. The steps to be performed are:

1. Load the MIB and generate the SNMP master agent. The code generated for proxy table remote configuration in the generated main file is:
   

   //Code generated will be as follows

   dyn = new DynamicRegistration(false, "conf", "ProxyTable.xml");
   dyn.addRegistrationListener(hdlr, true);

2. Change the boolean from 'true' to 'false' in the code snippet
   

   //Code to be changed

   dyn = new DynamicRegistration(false, "conf", "ProxyTable.xml");
   dyn.addRegistrationListener(hdlr, false);

3. Compile and run the master agent.

Questions

10. Can I add sub-agent to master agent at runtime without bringing other
agents down?

Yes. To achieve this, add sub-agent entries to the proxy table of the master agent by loading AGENT-SNMP-CONFIG-MIB packaged with the product. Please refer to the SNMPv1/v2 demo that illustrates the steps involved, while adding an WebNMS SNMP agent to a master agent.

Questions

11. Does the toolkit support MIB II implementation?

Yes. SNMP Agent Toolkit supports MIB II implementation through master-subagent architecture. WebNMS's SNMP agent can proxy other third-party SNMP agent supporting MIB II, by configuring it as sub-agent to WebNMS's SNMP agent.

Questions