monitor graph

Unanswered Question
May 29th, 2010
User Badges:

Hello ,


Although this is non-cisco , wanted to get an opinion on my query from folks here.


Attached is an image from mrtg output of a link. It has been observed that for past few days, the graph levels out as seen in the image file.

Anyone, please help me understand what could be causing this. Is this normal or not?


Appreciate all assistance.

Attachment: 
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (1 ratings)
Loading.
Joe Clarke Sun, 05/30/2010 - 10:57
User Badges:
  • Cisco Employee,
  • Hall of Fame,

    Founding Member

That's hard to say.  Normal is very subjective.  You should look at the octet counters on this interface to see if they match the pattern you are seeing in this graph.  It could very well be that this interface has reached a constant steady-state of traffic.  MRTG will calculate interface throughput by looking at the input and output octets over a five minute period.  The function is (for output bps):


delta(ifOutOctets) * 8

---------------------- = bps

         300

suthomas1 Sat, 06/12/2010 - 22:00
User Badges:

Thanks for your answer. Few more queries though, setup is as follows:


   Two routers at headend in HSRP mode ( loadsharing mode by bgp) , connected to two distribution mode layer 2 switches , which then go to different spoke sites.


With this setup, we are trying to gather usage for each spoke site. Both routers are configured as multiple encapsulation to the access switch ports for each spoke. we ran mrtg on both routers . Strange thing noticed is router1's graph looks near accurate whereas router2's graphs for spokes looks inaccurate with usage being very low and getting flat at bps.( which is highly unlikely ).


Please help as to why is this happening so? router2's graph should also be getting its stats from Switch2 as that is the access part.

Anyways to rectify this.


Thanks in advance!

Joe Clarke Sat, 06/12/2010 - 22:19
User Badges:
  • Cisco Employee,
  • Hall of Fame,

    Founding Member

What do the "show interface" counters report for the same interfaces being managed by MRTG?  Do they agree?

suthomas1 Sun, 06/13/2010 - 08:44
User Badges:

Corresponding interface counters on devices dont match with what is seen on graph. Interface shows quite good amount of traffic.




Thanks!

Joe Clarke Mon, 06/14/2010 - 22:58
User Badges:
  • Cisco Employee,
  • Hall of Fame,

    Founding Member

What kind of interface is this?  Post the output of "show interface" run twice with a five minute gap between each execution.  Additionally, post the output of an SNMP Walk of the ifTable taken at the same times as the "show interface" output.

suthomas1 Tue, 06/15/2010 - 09:30
User Badges:

Thanks. How do i get the snmp walk of iftable . Is there a command to get it, please let me know .


Thanks!

Joe Clarke Wed, 06/16/2010 - 23:06
User Badges:
  • Cisco Employee,
  • Hall of Fame,

    Founding Member

You will need an snmpwalk tool like the one provided in the Net-SNMP suite (http://net-snmp.sf.net).  Then, run the command:


snmpwalk -v 2c -c COMMUNITY IP ifTable

Actions

This Discussion