Advertisement
The SNMP system includes the following parts:
SNMP is a protocol working on the application layer. It provides the packet format between SNMP management side and agent.
SNMP management side can be part of the network management system (NMS, like CiscoWorks). Agent and MIB are stored on the system. You need to define the relationship between network management side and agent before configuring SNMP on the system.
SNMP agent contains MIB variables-SNMP management side can check or modify value of these variables. The management side can get the variable value from agent or stores the variable value to agent. The agent collects data from MIB. MIB is the database of device parameter and network data. The agent also can respond to the loading of the management side or the request to configure data. SNMP agent can send trap to the management side. Trap sends alarm information to NMS indicating a certain condition of the network. Trap can point out improper user authentication, restart, link layer state (enable or disable), close of TCP connection, lose of the connection to adjacent systems or other important events.
When some special events occur, the system will send 'inform' to SNMP management side. For example, when the agent system detects an abnormal condition, it will send information to the management side.
SNMP notification can be treated as trap or inform request to send. Since the receiving side doesn't send any reply when receiving a trap, this leads to the receiving side cannot be sure that the trap has been received. Therefore the trap is not reliable. In comparison, SNMP management side that receives "inform request" uses PDU that SNMP echoes as the reply for this information. If no "inform request" is received on the management side, no echo will be sent. If the receiving side doesn't send any reply, then you can resend the "inform request". Then notifications can reach their destination.
Since inform requests are more reliable, they consume more resources of the system and network. The trap will be discarded when it is sent. The "inform request" has to be stored in the memory until the echo is received or the request timeouts. In addition, the trap is sent only once, while the "inform request" can be resent for many times. Resending "inform request" adds to network communications and causes more load on network. Therefore, trap and inform request provide balance between reliability and resource. If SNMP management side needs receiving every notification greatly, then the "inform request" can be used. If you give priority to the communication amount of the network and there is no need to receive every notification, then trap can be used.
This OLT only supports trap, but we provide the extension for "inform request".
System of our company supports the following SNMP versions:
Layer 3 switch of our company also supports the following SNMP:
SNMPv1 uses group-based security format. Use IP address access control list and password to define the management side group that can access to agent MIB.
SNMPv3 provides secure access to devices by a combination of authenticating and encrypting packets over the network.
The security features provided in SNMPv3 are:
SNMPv3 provides for both security models and security levels. A security model is an authentication strategy that is set up for a user and the group in which the user resides. A security level is the permitted level of security within a security model. SNMPv3 supports three security levels based on the user's security model, that is (from high to low), authentication and encryption, authentication and no encryption, no authentication. With MD5 or SHA hash algorithm, the password will not be revealed. With DES encryption, the device will not be wiretapped by a third party. To realize identity authentication of the device, you need to configure user/password pair and the group belongs to the user. To determine the access permission to the management information database, you need to configure group and view. Meanwhile, the group limits the lowest security level of users in the group.
You need to configure SNMP agent to the SNMP version that the management working station supports. The agent can communicate with many management sides.
SNMP of our system supports all MIBII variables (which will be discussed in RFC 1213) and SNMP traps (which will be discussed in RFC 1215).
Our system provides its own MIB extension for each system.
SNMP Configuration Tasks include:
Command | Usage Guidelines |
snmp-server view name oid [excluded I included] | Adds the subtree or table of OID-specified MIB to the name of the SNMP view, and specifies the access right of the object identifier in the name of the SNMB view. |
The subsets that can be accessed in the SNMP view are the remaining objects that "include" MIB objects are divided by "exclude" objects. The objects that are not configured are not accessible by default.
After configuring the SNMP view, you can implement SNMP view to the configuration of the SNMP group name, limiting the subsets of the objects that the group name can access.
Command | Purpose |
snmp-server community [0|7] string [view view-name] [ro I rw] [word] | Defines the group access character string |
You can configure one or multiple group character strings. Run command "no snmp-server community" to remove the specified community character string.
For how to configure the community character string, refer to the part "SNMP Commands".
Command | Purpose |
snmp-server contact text | Sets the character string for the linkman of the node. |
snmp-server location text | Sets the character string for the node location. |
Command | Purpose |
snmp-server packetsize byte-count | Sets the maximum length of the data packet. |
Command | Purpose |
show snmp | Monitoring SNMP state |
Command | Purpose |
snmp-server enginelD local enginelD | Configuring SNMP local engine |
Command | Purpose |
snmp-server host|hostv6 host community-string [trap-type] | Specifies the receiver of the trap message |
snmp-server host|hostv6 host [vrf word] [udp-port port-num] [permit|deny event-id] {{version [v1 | v2c | v3]} I {[informs | traps] [auth | noauth]}} community-string/user [authentication | configure | snmp] | Specifies the receiver, version number and username of the trap message. |
When the system is started, the SNMP agent will automatically run. All types of traps are activated. You can use the command snmp-server host to specify which host will receive which kind of trap.
Some traps need to be controlled through other commands. For example, if you want SNMP link traps to be sent when an interface is opened or closed, you need to run snmp trap link-status in interface configuration mode to activate link traps. To close these traps, run the interface configuration command snmp trap link-stat.
You have to configure the command snmp-server host for the host to receive the traps
Command | Purpose |
snmp-server trap-source interface | Specifies the source interface where traps originate and sets the source IP address for the message. The command sets the source IP address for the information. |
snmp-server queue-length length | Creates the queue length of the message for each host that has traps. The default value is 10. |
snmp-server trap-timeout seconds | Defines the frequency to resend traps in the resending queue. The default value is 30 seconds. |
Command | Purpose |
snmp source-addr ipaddress | Set the source address for the SNMP message |
Command | Purpose |
snmp-server udp-port portnum | Set SNMP server udp-port number |
Command | Purpose |
snmp-server group [groupname { v3 [auth I noauth I priv]}][read readview][write writeview] [notify notifyview] [access access-list] | Configuring SNMPv3 group You can only read all items in the subtree of the Internet by default. |
Command | Purpose |
snmp-server user username groupname {v3 [encrypted|auth] [md5|sha] auth-password} | Configures a local SNMPv3 user. |
Command | Purpose |
snmp-server encryption | This command is used to display the SNMP community, the SHA encryption password and the MD5 encryption password in plain text. In this way, the security of the password is guaranteed. |
Command | Purpose |
snmp-server trap-source interface | When the SNMP server sends out a SNMP trap on whichever interface, the SNMP trap shall carry a trap address. If you want to use the trap address for tracking, you can use this command. |
Command | Purpose |
snmp-server trap-timeout seconds | Before switch software tries to send traps, it is used to look for the route of destination address. If no routes exists, traps will be saved in the retransmission queue. The server trap-timeout command decides the retransmission interval. |
Command | Purpose |
snmp-server trap-add-hostname | This command is a great help in some cases when the NMS needs to locate which host sends these traps. |
Command | Purpose |
snmp-server trap-logs | After this function is enabled, the trap transmission records of a device can be sent to the log server and then you can know more about the running state of the device. |
Command | Purpose |
snmp-server set-snmp-dos-max retry times | After this function is enabled, the trap transmission records of a device can be sent to the log server and then you can know more about the running state of the device. |
The command must be used with snmp-server host.
Command | Purpose |
snmp-server keep-alive times | Send keep-alive times regularly to the trap host. |
Command | Purpose |
snmp-server necode text | The command is corresponding to the snmp private MIB variable |
Command | Purpose |
snmp-server event-id number trap-oid oid | The command is used to forward the filter when sending trap in configuring host. |
Command | Purpose |
snmp-server getbulk-timeout seconds | The command is used to set the timeout of processing getbulk request. If all getbulk requests cannot be processed in timeout, the system will return to the current result directly. |
Command | Purpose |
snmp-server getbulk-delay ticks | The command is used to set getbulk-delay time to prevent snmp from occupying excessive cpu when snmp agent processing getbulk request. The unit is 0.01s |
Command | Purpose |
show snmp enginelD | Shows SNMP engine information. |
show snmp host | Shows SNMP trap host information |
show snmp view | Shows SNMP view information |
show snmp mibs | Shows SNMP MIB registration information |
show snmp group | Shows SNMP group information |
show snmp user | Shows SNMP user information. |
Command | Purpose |
debug snmp error | Enable the debug OLT of SNMP error information |
debug snmp event | Enable the debug OLT of SNMP event information. |
debug snmp packet | Enable the debug OLT of SNMP input/output packets |
RMON configuration tasks include:
You can configure the rMon alarm function through the command line or SNMP NMS. If you configure through SNMP NMS, you need to configure the SNMP of the switch. After the alarm function is configured, the device can monitor some statistic value in the system. The following table shows how to set the rMon alarm function:
Command | Purpose |
config | Enters the global configuration mode |
rmon alarm index variable interval {absolute | delta} rising-threshold value [eventnumber] falling-threshold value [eventnumber] [owner string] [repeat] | Add a rMon alarm item. index is the index of the alarm item. Its effective range is from 1 to 65535. variable is the object in the monitored MIB. It must be an effective MIB object in the system. Only objects in the Integer, Counter, Gauge or TimeTicks type can be detected. interval is the time section for sampling. Its unit is second Its effective value is from 1 to 2147483647. absolute is used to directly monitor the value of MIB object delta is used to monitor the value change of the MIB objects between two sampling. value is the threshold value when an alarm is generated. Event number is the index of an event that is generated when a threshold is reached. Event number is optional. Owner string is to describe the information about the alarm Repeat is to repeat trigger event. |
exit | Goes back to the EXEC mode |
write | Saves the settings |
After a rMon alarm item is configured, the device will obtain the value of variable-specified oid after an interval. The obtained value will be compared with the previous value according to the alrm type (absolute or delta). If the obtained value is bigger than the previous value and surpasses the threshold value specified by rising-threshold, an event whose index is eventnumber (If the value of eventnumber is 0 or the event whose index is eventnumber does not exist in the event table, the event will not occur). If the variable-specified oid cannot be obtained, the state of the alarm item in this line is set to invalid. If you run rmon alarm many times to configure alarm items with the same index, only the last configuration is effective. You can run no rmon alarm index to cancel alarm items whose indexes are index.
The steps to configure the rMon event are shown in the following table:
Procedure | Command | Purpose |
1 | config | Enters the global configuration mode |
2 | rmon event index [description string] [log] [owner string] [trap community] [ifctrl interface] | Add a rMon event item. index is the index of the alarm item. Its effective range is from 1 to 65535. description means the information about the event. log means to add a piece of information to the log table when a event is triggered. trap means a trap message is generated when the event is triggered. community means the name of a community. ifctrl interface is the interface controlling event shutdown. owner string is to describe the information about the alarm. |
3 | exit | Goes back to the EXEC mode. |
4 | write | Saves the settings. |
After a rMon event is configured, you must set the domain eventLastTimeSent of the rMon event item to sysUpTime when a rMon alarm is triggered. If the log attribute is set to the rMon event, a message is added to the log table. If the trap attribute is set to the rMon event, a trap message is sent out in name of community. If you run rmon event many times to configure event items with the same index, only the last configuration is effective. You can run no rmon event index to cancel event items whose indexes are index.
The rMon statistics group is used to monitor the statistics information on every port of the device. The steps to configure the rMon statistics are as follows:
Procedure | Command | Purpose |
1 | config | Enters the global configuration mode |
2 | interface iftype ifid | This command is used to enter the interface configuration mode. iftype means the type of the port. ifid means the ID of the interface. |
3 | rmon collection stats index [owner string] | Enable the statistics function on the port. index means the index of the statistics. owner string is to describe the information about the statistics. |
4 | exit | Goes back to the global mode. |
5 | exit | Goes back to the EXEC mode. |
6 | write | Saves the settings. |
If you run rmon event many times to configure status items with the same index, only the last configuration is effective. You can run no rmon event index to cancel event items whose indexes are index.
The RMON history group is used to collect statistics information of different time sections on a port in a device. The steps to configure the rMon statistics are as follows:
Procedure | Command | Purpose |
1 | config | Enters the global configuration mode. |
2 | interface iftype ifid | Enters the port mode. iftype means the type of the port. ifid means the ID of the interface. |
3 | rmon collection history index [buckets bucket-number] [interval second] [owner owner-name] | Enable the history function on the port. index means the index of the history. In statistics of all history record control entries, the entry nearest to bucket-number needs to be saved. The user can browse the Ethernet history record to obtain the statistics The default value is 50 entries. The interval means the time between two data collection, whose default value is 1800s (half hours). owner string is to describe the information about the description information in the history control table. |
4 | exit | Goes back to the global mode |
5 | exit | Goes back to the EXEC mode. |
6 | write | Saving the Settings |
After a rMon history item is added, the device will obtain statistics values from the specified port every second. The statistics value will be added to the history item as a piece of information. If you run rmon collection history index many times to configure history items with the same index, only the last configuration is effective. You can run no rmon history index to cancel history items whose indexes are index. Note: Too much system sources will be occupied in the case the value of bucket-number is too big or the value of interval second is too small.
Run show to display the RMON configuration of the switch.
Command | Purpose |
show rmon [alarm] [event] [statistics] [history] | Displays the rmon configuration information. alarm means to display the configuration of the alarm item. event means to show the configuration of the event item and to show the items that are generated by the occurrence of events and are contained in the log table. statistics means to display the configuration of the statistics item and statistics values that the device collects from the port. history means to display the configuration of the history item and statistics values that the device collects in the latest specified intervals from the port. |
Here you can download full pdf version of manual, it may contain additional safety instructions, warranty information, FCC rules, etc.
Download FS S3150-8T2FP, S3260-8T2FP, S3260-16T4FP, PoE+ Series Manual
Advertisement
Need help?
Do you have a question about the PoE+ Series and is the answer not in the manual?
Questions and answers