LMS 3.0 Important

Answered Question
May 23rd, 2008

1-I have remote sites and main branch connection between remote sites to main branch is MPLS

2-Connection between main branch to remote sites over 2 ISP

My question

I need real time monitor with out enable GRE tunnel between main branch to remote site

If the router down in remote site from any ISP I need the router will be reed

I need this action in topology map

I don't need this action by mail or by syslog

I have this problem too.
0 votes
Correct Answer by Joe Clarke about 8 years 7 months ago

DFM monitors all devices in a Known state. If the devices are healthy (e.g. they are reachable, and have no down interfaces) then you will not see any alert for them in AAD. DFM is still managing and monitor those devices, though.

Correct Answer by Joe Clarke about 8 years 8 months ago

AAD only shows alerts. Devices which have no faults will not be shown in AAD. You can get a list of all devices DFM is managing under DFM > Device Management > Device Status. Any device in a Known state is being properly managed by DFM.

Correct Answer by Joe Clarke about 8 years 8 months ago

AAD automatically refreshes every 30 seconds. This cannot be customized.

Correct Answer by Joe Clarke about 8 years 8 months ago

AAD = Alerts and Activities Display

You can filter the output on the fly by clicking on the filter button at the top of the AAD screen. You can create custom filter groups by going to DFM > Configuration > Other Configurations > Alerts and Activities Defaults, and creating groups there.

Correct Answer by Joe Clarke about 8 years 8 months ago

The AAD will show alerts for ALL devices. An alert will appear for each device having a fault. The AAD can be filtered on device type, device names, and alert IDs. DFM, unlike the rest of LMS, has limitations on the number of supported ports and interfaces. If you only wanted to manage the two serial interfaces on 1300 devices (i.e. 2600 interfaces), you could do that all with one instance of DFM.

Correct Answer by Joe Clarke about 8 years 8 months ago

The Alerts and Activities Display is designed to be monitored by operators. It automatically refreshes. The polling interval is customizable, but by default, it is four minutes.

I do not understand the second question.

Correct Answer by Joe Clarke about 8 years 8 months ago

It depends on the interface type, but in general DFM manages all interfaces by default. To confirm if your interfaces are managed, go to DFM > Device Management > Device Details, and bring up the Detailed Device View on your devices. Check to make sure the managed state of your interfaces is true. If not, set the state to true, then go to DFM > Configuration > Polling and Thresholds > Apply Changes to apply the changes.

After that, DFM will poll the interfaces to make sure they are up. If they go down, you will get an alert under DFM > Alerts and Activities > Alerts and Activities. Within the Alert you will have an OperationallyDown event pertaining to each down interface.

Correct Answer by Joe Clarke about 8 years 8 months ago

If a device becomes SNMP unreachable, then Campus will detect that on its next Data Collection or polling pass. Campus Manager is NOT a fault management application, and its map is not updated in real time.

By default, the polling interval is every two hours. Campus 5.0 and higher has a critical devices poller which runs every five minutes by default. After that update runs, you may have to manually refresh the topology map. The N-Hop portlet should update based on your refresh interval.

If you see devices not getting updated with their status, make sure that they are actually becoming SNMP unreachable when Campus runs its Data Collection or polling cycle. If so, enable frontend, framework, core, and corex debugging under Campus Manager > Admin > Debugging Options > Data Collection. Wait for a new polling cycle or Data Collection to run, then look at the ani.log for errors.

All that said, it would be better to use a more real-time fault management application like DFM, HPOV Network Node Manager, NetView, etc. to let you know when this branch site goes offline.

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (10 ratings)
Loading.
Correct Answer
Joe Clarke Fri, 05/23/2008 - 11:32

If a device becomes SNMP unreachable, then Campus will detect that on its next Data Collection or polling pass. Campus Manager is NOT a fault management application, and its map is not updated in real time.

By default, the polling interval is every two hours. Campus 5.0 and higher has a critical devices poller which runs every five minutes by default. After that update runs, you may have to manually refresh the topology map. The N-Hop portlet should update based on your refresh interval.

If you see devices not getting updated with their status, make sure that they are actually becoming SNMP unreachable when Campus runs its Data Collection or polling cycle. If so, enable frontend, framework, core, and corex debugging under Campus Manager > Admin > Debugging Options > Data Collection. Wait for a new polling cycle or Data Collection to run, then look at the ani.log for errors.

All that said, it would be better to use a more real-time fault management application like DFM, HPOV Network Node Manager, NetView, etc. to let you know when this branch site goes offline.

ahmed_saleh Fri, 05/23/2008 - 12:02

1-Please can you tell me all things about uses topology map?

2-if I put all network 1300 device in critical poler and make it every 2 sec is this wrong I need advice

Joe Clarke Fri, 05/23/2008 - 12:06

The criticial devices poller is designed for no more than 30 devices. You will cripple Campus if you drop the poller and add that many devices.

The main use for the Topology Map is network documentation. That is, use it for visualizing layer 2 topology, spanning-tree forwarding and blocking ports, traffic patterns (if using RMON data), and potential misconfigurations in the network.

ahmed_saleh Fri, 05/23/2008 - 12:41

Thanks for your help

Now I need ask about if I can configure DFM to monitor 2 serial interfaces for all remote branches and make real time monitor report in screen

Correct Answer
Joe Clarke Fri, 05/23/2008 - 12:47

It depends on the interface type, but in general DFM manages all interfaces by default. To confirm if your interfaces are managed, go to DFM > Device Management > Device Details, and bring up the Detailed Device View on your devices. Check to make sure the managed state of your interfaces is true. If not, set the state to true, then go to DFM > Configuration > Polling and Thresholds > Apply Changes to apply the changes.

After that, DFM will poll the interfaces to make sure they are up. If they go down, you will get an alert under DFM > Alerts and Activities > Alerts and Activities. Within the Alert you will have an OperationallyDown event pertaining to each down interface.

ahmed_saleh Fri, 05/23/2008 - 13:00

I can monitor this report with out generate or refresh I mean real time monitor

And can I customize this view for 2 cloud 192.18.0.0& 192.29.0.0

Correct Answer
Joe Clarke Fri, 05/23/2008 - 13:04

The Alerts and Activities Display is designed to be monitored by operators. It automatically refreshes. The polling interval is customizable, but by default, it is four minutes.

I do not understand the second question.

ahmed_saleh Fri, 05/23/2008 - 13:13

I need to ask about I have 2 serial interfaces in all branches to connect for main branch one serial have 172.18.0.0 & second serial have 172.29.0.0 can I customize report real time monitor for this interfaces

Correct Answer
Joe Clarke Fri, 05/23/2008 - 13:22

The AAD will show alerts for ALL devices. An alert will appear for each device having a fault. The AAD can be filtered on device type, device names, and alert IDs. DFM, unlike the rest of LMS, has limitations on the number of supported ports and interfaces. If you only wanted to manage the two serial interfaces on 1300 devices (i.e. 2600 interfaces), you could do that all with one instance of DFM.

Correct Answer
Joe Clarke Fri, 05/23/2008 - 13:27

AAD = Alerts and Activities Display

You can filter the output on the fly by clicking on the filter button at the top of the AAD screen. You can create custom filter groups by going to DFM > Configuration > Other Configurations > Alerts and Activities Defaults, and creating groups there.

ahmed_saleh Sun, 05/25/2008 - 04:07

Dear sir if I need customize this report Alerts and Activities Display to be refreshing every 30 sec not refresh after 4 mints please tell me the steps

Correct Answer
Joe Clarke Sun, 05/25/2008 - 08:28

AAD automatically refreshes every 30 seconds. This cannot be customized.

ahmed_saleh Sun, 05/25/2008 - 08:39

Now in DFM I can see the fault devices where I can see the normal devices

Correct Answer
Joe Clarke Sun, 05/25/2008 - 08:42

AAD only shows alerts. Devices which have no faults will not be shown in AAD. You can get a list of all devices DFM is managing under DFM > Device Management > Device Status. Any device in a Known state is being properly managed by DFM.

ahmed_saleh Sun, 05/25/2008 - 08:48

Ok but I have problem with DFM some alerts come with operation down when I try ping this device I get response can DFM when get alert with interface down after interface is up can DFM send alert to tell me the interface is up

Joe Clarke Sun, 05/25/2008 - 08:54

Alerts are composite entities. They are composed of one or more events. An event is an atomic problem on a device. An OperationallyDown event will clear automatically when DFM detects that the device or interface is once again reachable. Once all of the events within an alert clear, then the alert itself will clear.

You will see a cleared status in AAD for 30 minutes before the alert is removed from AAD. At such time, you will only be able to see it in Fault History for the next month.

It's important to note that DFM will treat any managed interface which is administratively up but operationally down as being OperationallyDown. If you have such interfaces on your devices that should be down, then be sure to "shutdown" those interfaces. DFM also pings each IP address on a device for reachability. If all the IP addresses on a given device are not ICMP reachable, then you will need to go to DFM > Device Management > Device Details, and unmanage the unreachable IP addresses. This is only doable in DFM 3.0 and higher. Note: you CANNOT unmanage the IP address by which you added the device to DFM or the device will go into a Questioned state.

ahmed_saleh Mon, 05/26/2008 - 07:17

Hi Joe

Please I attached file please I need you tell me why this event not clear from AAD

And please if you can explain some of this event

And thank for your help

Attachment: 
Joe Clarke Mon, 05/26/2008 - 09:40

I can't tell you why these are not clearing since I don't know what the underlying events are, nor do I know what is happening in your network. But, as I said in my previous post, alerts will clear when ALL of the underlying events have cleared. So go through each alerts' events, and see why they are being triggered, then either correct the problem on the device, or change your threshold setting.

ahmed_saleh Mon, 05/26/2008 - 23:32

Dear Joe

I need to ask you about some point

1- Can I use hosts file instead DNS and I need example for 1 router have 5 interfaces I will put all interfaces or I will put one interface in this hosts file if I can use hosts file

2-in topology we have MPLS between branches and can't enable JRE tunnel you have any solution to get link between them

3- can I discover node with out use loop back interface and I need to know what happen if I put ( example router have 2 serial interface I will enter the 2 interfaces in discovery like 172.18.1.1 &172.29.1.1)

4-can Cisco works help me to obtain monitoring screen to monitor the devices and to display the status of the device when it becomes up or down real time monitor I need up and down I don't need use AAD

Joe Clarke Mon, 05/26/2008 - 23:48

1. Yes. You can add the one interface by which you manage this device into the hosts file.

2. If you cannot get CDP on a link between two campuses, then you will not get a link drawn on the Topology Map. There is no workaround for that.

3. Discovery will eventually fall back to serial interfaces if you select the use loopback option. You don't need to multiple seed device entries per device. Really, you only need one seed device per campus, or one seed device per disjoint area of your network.

4. If you don't want to use DFM's Alerts and Activities Display, then there is nothing in LMS that will give you fault information for devices in a real-time view.

ahmed_saleh Tue, 05/27/2008 - 01:47

Hi Joe

Now I can't display in AAD the up devices and I have problem in network because MPLS all site obtain as separate

1-Can I add 1 router in CS by 2 serial interfaces to obtain in topology map as 2 routers?

2-if you have any solution I can use to real time monitor to up and down interfaces please tell me

Joe Clarke Tue, 05/27/2008 - 08:30

You can add the same device twice in DCR provided each of the hostname, display name, and IP are different. However, applications like RME and Campus will detect the device is the same. Campus won't duplicate the device unless there is a problem.

Unlike a graphical fault management application, DFM doesn't show green, "all-is-well" icons to indicate a device is healthy. You just don't see any alerts from a healthy device in AAD (or the status of an alert will be Cleared indicating a problem has been resolved).

The upcoming LMS 3.1 release will offer integration between DFM and Campus, so there will be some visual indication of faults on the topology map, but that is not scheduled until the end of next month.

ahmed_saleh Tue, 05/27/2008 - 10:09

Sorry joe for all these question but I need know because customer when ask me

1-I run automatic discovery and I put seed device with 2 serial interfaces for each remote site and enable loopback interface why in topology the device appear with ip for loopback interface

2- DFM for fault devices its ok where I can monitor normal devises this question customer ask me I need answer to tell hem

Joe Clarke Tue, 05/27/2008 - 10:13

If you selected to use loopback when determining the management IP, then that's what Discovery will do. It will add the device to DCR with the loopback IP. Campus will then use that IP to manage the device.

I don't understand the second question.

Correct Answer
Joe Clarke Tue, 05/27/2008 - 11:19

DFM monitors all devices in a Known state. If the devices are healthy (e.g. they are reachable, and have no down interfaces) then you will not see any alert for them in AAD. DFM is still managing and monitor those devices, though.

ahmed_saleh Tue, 05/27/2008 - 12:02

Joe

If core switch have sub interfaces connected to remote sites. And I will but this core switch in critical poller if any router down in remote sites the link between sub interfaces from core switch to router in remote sites will be down in topology map

Joe Clarke Tue, 05/27/2008 - 12:27

If a link that exists on the topology map goes down according to SNMP (i.e. the ifOperStatus is down), then the topology map should reflect this after the next polling cycle.

ahmed_saleh Wed, 05/28/2008 - 09:24

Joe

1- I take for you screen shot for problem in DFM the bandwidth is 2M but ciscoworks get the bandwidth 256 and get me alert about utilization why .these problem form 14 day tell know

2- I make hosts file and put it in C:\WINDOWS\system32\drivers\etc and restart the server and run discovery and run data collection but the device not appear with name

Joe Clarke Wed, 05/28/2008 - 09:32

You should check ifSpeed for this interface to make sure it is advertising the 2 Mbps value.

You will need to remove and re-add devices from DFM to get the name updates.

ahmed_saleh Wed, 05/28/2008 - 10:35

1-How I can check ifSpeed I need steps

2-You will need to remove and re-add devices from DFM to get the name updates.

This Is will help me in all application like CM, RME, IP or I will remove devices from all application

ahmed_saleh Wed, 05/28/2008 - 11:31

How I can enable cdp over mpls between to site to discover this site by Cisco works

Joe Clarke Wed, 05/28/2008 - 11:34

I do not believe this is possible. However, the typical way to enable CDP on a non-broadcast interface is to add the command "cdp enable".

Joe Clarke Wed, 05/28/2008 - 11:33

Perform an snmpwalk of ifSpeed on this device. Then perform an snmpwalk for ifDescr. Match the ifIndex values to confirm the ifSpeed value.

No, this will help for DFM. For other applications, you will need to modify the devices' display names in DCR, or configure Discovery to "Update DCR Display Name" under Common Services > Device Discovery > Discovery Settings > Global Settings. Then you will need to run another Discovery.

ahmed_saleh Wed, 05/28/2008 - 21:43

Joe I don't understand about how I can Perform an snmpwalk of ifSpeed on this device . did you mean I will enter to device and Right snmpwalk

And I don't under stand this also Then perform an snmpwalk for ifDescr how I can make this

ahmed_saleh Thu, 05/29/2008 - 02:16

I am enabled CDP on 6500 on interface POS

And enabled CDP on router in remote site in serial interfaces but I can't get the link between main branch and remote site in topology map

Attachment: 
Joe Clarke Thu, 05/29/2008 - 08:54

What does show cdp neighbor report on each end of the link? Just because you can enable CDP doesn't mean it will actually work.

ahmed_saleh Thu, 05/29/2008 - 12:21

What does show cdp neighbor report on each end of the link?

In branch just I see direct connected device from my network put in wan interface I see the router from ISP

In remote site I see the connected switch from my site put in wan serial interface I didn't see any thing

Joe please tell me what can do. To do I need Cisco works drown the link I need solution

Joe Clarke Thu, 05/29/2008 - 13:34

Given devices A and B, for Campus Manager to show a link between these two devices, show cdp neighbor on A must show B, and show cdp neighbor on B must show A. If that is not happening, you either need to fix it, or you will not see a link between these devices. Typically, MPLS clouds cannot have links on the topology map.

ahmed_saleh Thu, 05/29/2008 - 15:19

Yes I show the link between 2 core switches its ok in topology but I can't see the remote site links in topology how I can fix it

Yes I can't show main branch from remote site when I write command show CDP

And I know CDP can get direct connected device but here I have the MPLS is direct connected what is the solution now to resolve this problem to get drown links in topology map

Joe Cisco designed Cisco works to work over LAN or what. Cisco works can't get links over wan how is this. Tell know you didn't tell me what I do I am from Cisco partner (MIS) and I am install Cisco works for customer tell me what I do to do if Cisco works cant get links tell me to told the customer or resolve this point or tell me solution to get links drown in topology map

Joe Clarke Thu, 05/29/2008 - 08:50

You can use the SNMP Walk tool under Device Center > Tools > SNMP Walk to do this. Just specify the starting OID as ifSpeed then ifDescr.

ahmed_saleh Thu, 05/29/2008 - 12:23

What did you mean about? Just specify the starting OID as ifSpeed then ifDescr I need more clarify

Joe Clarke Thu, 05/29/2008 - 13:35

For the first walk, specify ifDescr as the starting OID. For the second walk, specify ifSpeed as the starting OID.

ahmed_saleh Thu, 05/29/2008 - 15:26

Joe

I need steps and example I don't under stand tell know

What did you mean about? (Specify ifDescr as the starting OID) and where I write it

What did you mean about? (Specify ifSpeed as the starting OID) and where I write it

Give me an example and tell me the steps

Actions

This Discussion