for network management are the following operations to be performed on a network: configuring, monitoring, throubleshooting.
SNMP protocol is an example of management and is the most commonly used within an IP network. It consists essentially of three components: -A manager is the devices (PC or router) is used to display the result of monitoring. You can use several tools for displaying results, such as HP-Opem View
an agent is the device (access servers, routers, switches) to monitor
-A protocol used for communication between agent and server (SNMP)
Funzionamemto:
are connected via UDP port 161 and 162 for the agent to the Manager. The manager, when they want to require the info and info is being transmitted between agents and managers through the use of SNMP.
addition to this pool with some frequency by the manger, there is the possibility to activate a trap or information sent by the agent directly to the manager, for example when some event is triggered.
SNMP Community string is a password set on the agent, that allows the manager to access ("Access-Level") to the MIB according to different modaltà :
-Read-only-Read-write: the manager has full access to the MIB, but can not change the community string
-Read-write -all: The manager can do everything (read, write mib, change community string)
configuration steps:
1) the community string is set on the agent and the associated access-lievel "
S ( config) # snmp-server community ro readmeCommunity S (config) # snmp-server community rw writemeCommunity
2) Can I enable the agent to inform the manager of information, for example, if something happens:
S (config ) # snmp-server trap {enable through the MIB browser, the manager has the ability to query on the agent to view the information.
For example you can use the following MIBs: MIB
Tree.router_std MIBs.iso.org.dod.internet.mgmt.mib-2.system.sysDescr
(ref: http://www.webnms.com / cagent / help / technology_used / c_snmp_overview.html # mib)
to see, for example, the IOS version on router "Medway, ME."
In this particular case, the manager sends a request of Get to the agent, and is then able, through MIB Browser displays the value or the request sent.
Reference Cisco http://www.cisco.com/public/sw-center/netmgmt/cmtk/mibs.shtml
Example with GNS3
Here is a LAB that is used:
- the agent is a router 3640, which IOS is c3640-mz.124-16a-jk9o3s
- while the manager is installed MibBrowser on my local PC
- clearly there is full visibility between the router and my PC
Step 1: Choose the
Cisco MIB provides MIB according to the device that you want to do the management.
For example, in our case, I have a 3640 with a particular IOS. From the site http://tools.cisco.com/ITDIT/MIBS/MainServlet?ReleaseSel=3407&PlatformSel=81&fsSel=0
you can select the MIB associated with your device, the version IOS and feature set supported. Once you select the MIB of interest, I have saved locally and then loaded on mibbrowser At this point you can query the agent using the Mib Browser.
The topology of reference is:
The interface of the router has the ip address 192.168.255.29, and then in the mib MibBrowser will call for pointing to this address.
IP SLA
Network Management Tool used for, or verify that the network is working properly for example.
With IP SLA is an agreement between the service provider and the company holding network, quality of service provided by the service provider and perceived by the user. The network administrator can monitor the network, and verify that the agreement is ripsettato ISP, or track to take proactive action on any problems of the network. IP SLA measures are: -jitter, latency, packet loss -RTT, Round Trip Time
IP SLA configuration implies the configuration of a source that monitors and another device configured as a responder who is "monitored".
The device that acts as IP SLA Source sends probe to which the responder must clearly respond. These probes are then used to calculate the measures deiderate (jitter, RTT, packet loss etc)
The responder can be qualiasi IP system, but if it is a Cisco device properly set as responders, the measure will be more accurate. Before starting the actual measurement of scmbio relished there is a source of info and IP SLA responder referred to as Control Protocol (UDP port 1967).
Configurations:
-Source:
Define the identifier ip-sla-
Define the operation and address of the target
-Define the frequency-
Define the measure when it starts -Define the possible reaction
Typically, the source collects the necessary info, stores it in the MIB which can then be read via SNMP. If
to monitor specific S3 S4 S3 and S4 should be monitored.
Solution Plan 1: S3 S4 monitors
Source-> Responder S3-> S4
Operation -> icmp echo, an operation used to measure the time interval between an echo from a source and a reply by any device into an IP address (in this case do not have to configure the responder) S3 (config) # ip sla 1
S3 (config-ip-sla) # icmp-echo "ip address of any interface of the switch S4"
/ / It's like if I did a ping interface
Solution Plan 2: S3 S4 monitors
Source-> Responder S4-> S3 S4 (config) # ip sla 1
S4 (config-ip-sla) # icmp-echo "ip address of any interface of the switch S3"
IP SLA Operation Before starting the monitoring, there is a kind of synchronization between source and responder, follow the steps listed below: 1) The source sends an IP SLAs control message indicating the operation that also want to use UDP port 1967. The control message contains the protocol, port, and the operation defined on the source router.
-MD5 is enabled even if the checksum is sent
-You can also enable authentication: if it fails the first destination responder sends a message authentication failaure
-If a response is not received by the responder, the source always tries to resend the message.
2) The responder sends a confirmation message when it receives the invitation and is set to listen on the specified port
3) If the answer is OK, the source starts to send probe packets
4) The responder responds
0 comments:
Post a Comment