Re: [nv-l] NVSNIFFER Capabilities

[Jeff Fitzwater <jfitz@Princeton.EDU>]

 

Scott Hammons wrote:

Listers,

First, thanks to all that replied to my earlier post.I?m still a little confused as to what my options are if I decide to use the nvsniffer ?feature?.

Can someone explain to me the capabilities of the nvsniffer option?I read the man page and it describes how you can create smartsets that contain the port information from monitored servers.Does this mean I can configure NetView to monitor specific TCP ports (ie., 94, 9494, etc?)If so what type of information does this provide?Is anyone using this feature?How does this compare to other sniffer applications?

I?m trying to come up with a plan to monitor the different TCP ports for the applications (see my earlier post).I would like to avoid introducing third party applications if at all possible; however I realize there are some limitations to what NetView can provide.Other than the man page, the NetView documentation is vague when it comes to this particular ?feature?.

Thanks in advance for the responses.

Scott

Scott the NVSNIFFER is not a packet sniffer as one would think, the tool just opens and closes the TCP ports you specify on the hosts or smartset you specify, and if the host has the port enabled (an application attached or just listening like the discard port) then nvsniffer knows that port is functional.  I believe the lowest polling resolution is 5 min.    You  can only do TCP ports as I suspect that is what you need.
 I would read the man page and doc on nvsniffer to understand the full capabilities if it.
Here at Princeton we use a tool called ROVER that reads a config file that has the hosts we want to monitor and also has what ports we want to test on each.   I had hoped to replace ROVERs functions with NVSNIFFER.   NOTE we acquired ROVER code many years ago when it was free and crude, and have since heavily modified it to suit Princeton.
 

Jeff Fitzwater
OIT Systems & Networking
Princeton University