I defined DefaultIpEcho collector between CAT65K (source) and GSS, CAT65K (source) and ASA. But in both cases the collector status is showing as 'configuring'.
What does it imply.
This could simply mean you need to force a refresh of the Collector Management screen. However, it could indicate a problem with the SNMP read-write credentials in DCR, configuration on the switch, or the software running on the switch.
CAT65K is running 12.2(33)SXH3 which support "ip sla monitor" commands.
Is SNMP read only string sufficient for IPM or is write string mandatory.
If I need to modify the device credentials and do the changes in DCR via Common Services would it automatically be picked by device listing in RME, IPM, HUM etc or is there a refresh mechanism for these modules.
You must have SNMP read-write to use IPM. Yes, as soon as you add new credentials to DCR, they will be picked up by the various applications.
I restart the Daemon Manager and then restarted the server itself. When I invoke IPM, I get the following error
Error in communicating with IPMProcess or IpmDbEngine.
It may be down or not yet up. Please make sure processes are up and running
Will I be able to use ASA as the target device because ASA's do not support SNMP write strings. Does IPM use read-write only for the IP SLA device (Source device).
SNMP is only required for the source device. The target devices can be any IP addressable node for the icmpEcho operation.
Ok. I will check on that.
What is the difference between operation types defaultIpecho, Echo connectionloss, Echo timeout, Echo rtt.
They all show Latency graph with the same value.
Also what is the difference between Echo and EthernetPing operation types.
Lastly, why doesn't EthernetPing show on the system-defined operation list but only user defined.
Are you referring to reaction types? There is only one icmpEcho. There are other built-in operations to adjust the QoS for the ICMP data, but everything else is the same. The reaction types allow you to control when a reaction event is triggered (e.g. when a trap is sent). This can be if a connection loss event is detected, if the collector exceeds the pre-configured timeout, or if the total round-trip response time is above a configured threshold. A reaction configuration is optional.
Echo refers to ICMP ping. Ethernet Ping is used in metro ethernet configurations to test layer 2 connectivity over ethernet.
The Ethernet Ping requires additional configured, and a completely configured pre-built operation cannot be created.
With regards to DCR if the management IP's are configured, all modules RME, Campus Manager etc would use them.
However, if the application performance is to be gauged then IPM should be configured on Production traffic Interface and not the Management interface/IPs.
How can we manage different set of IPs for IPM target device to check the network performance.
While IPM is capable of using a configurable source IP address, it is not yet capable of using a specific destination target address. I filed an enhancement request for this a while ago (CSCsz66719). However, a patch has been created to add this functionality. The patch (for IPM 4.2) is available from the TAC.
Yes, if possible, that will work. However, if you want to measure UDP jitter, then adhoc targets will not work.