DX Unified Infrastructure Management

I've been testing monitoring the network traffic of windows systems with the newest CDM probe, which added this functionality that is only configurable through the admin console.

I was wondering if anyone knows how this information is gathered.  I have previously been using the ntperf probes to collect this data for all NICs on any system.  For example I have a server with 2 NICs and get 2 metrics each for Total, Sent, and Received stats using ntperf.  Using the CDM probe, this data does not match up and is always higher than ntperf's data and there is only 1 metric per target even though there are 2 nics.  Even after adding up the two presented by ntperf.  I currently have both CDM and NTPERF checking these stats every minute.

Average (combined) from NTPERF for the last hour is: 12330 bytes/sec

Average from CDM for the last hour is: 16286 bytes/sec

Max (combined) from NTPerf: 18280 bytes/sec

Max from CDM: 33321 bytes/sec

Maybe it's just the 1 minute sample times falling at different intervals within that minute, but thought I'd ask here if anyone had insight.

Good question! I also wonder if this would require cdm to run as root on linux systems in order to get this information?

One can also wonder why they added this feature to the CDM probe, when they already have the net_traffic probe...

Hi,

As far as i am concerned, net_traffic is about to be deprecated... A pitty as it is a vers useful One.

integrating such feature in CDM could be a good idea...but so far the data collected aggregates all interfaces... And for now it looks like  the template performance reports linked to that data displayed in USM seems not working (we only tested it on migrating systems...not a full fresh UIM/UMP install)

GUillaume

Guillaume, I've had some trouble with the data being displayed in UMP as well. It's also funny that the "interfaces" and the default tab show metrics from different probes (although they could be from several of them). Also it is important that S_QOS_SNAPSHOT table is being updated timely by the data_engine. I bumped into a situation where this was not the case and I didn't see any metrics in USM for interfaces.

-jon