[mrtg] Strange MRTG problem

Wed May 15 21:19:21 CEST 2013

If you're running MRTG on 10 minute intervals, 
isn't the "Interval: 5" setting in your config file
wrong?  If the rrd file was created with Interval: 5,
and then you change the polling interval to 10, but
don't dump/recreate the rrd file, it'll still be operating
under the assumption your interval is 5 minutes, not
10 minutes.

Does your data start to look better if you run it at
five minute intervals, to match what your rrd file
is expecting?

Matt

>________________________________
> From: Simon Westlake <simon.westlake at digitalgunfire.com>
>To: 
>Cc: "mrtg at lists.oetiker.ch" <mrtg at lists.oetiker.ch> 
>Sent: Wednesday, May 15, 2013 4:31 AM
>Subject: Re: [mrtg] Strange MRTG problem
> 
>
>
>Thanks for the ideas. Community string is actually redacted, but I guess I should've made that more obvious than mashing the keyboard :)
>
>
>Agreed, my first thought was v1 but I am now specifically targetting 64bit counter OIDs and I've verified it is using SNMP v2 through debug options. Definitely sure the port is the same - low values are slightly different but the intervals are also slightly different (Cacti is running on 1 minute, MRTG is running on 10 minutes.)
>
>
>The MRTG version is super old - that was one of my first thoughts, I'm just somewhat reluctant to even touch it as it's running on an incredibly old version of Linux as well but I think that is my next plan.
>
>
>Not using a third party front end, it is a custom PHP script that calls 'rrdtool graph' directly. I will try rearranging the config file, thanks for that idea.
>
>
>
>On Tue, May 14, 2013 at 10:34 PM, Steve Shipway <s.shipway at auckland.ac.nz> wrote:
>
>A few thoughts
>> 
>>1.       You’ve just told the world your SNMP community string.  Oops.
>>2.       It certainly looks a bit like an SNMPv1 issue, though that would generally top off about 120Mbps.  However the config seems to use v2 correctly.
>>3.       Are you absolutely certain that both graphs relate to the same port?  Possibly there is some confusion in the settings?
>>4.       Look at the two graphs close up.  Are they exactly the same for low values, then lose it for high values?  If so, maybe its using SNMPv1 anyway.  If not, maybe you’re graphing something different.
>>5.       Which version of MRTG do you have?  Possibly an old one is not handling things correctly.  Get a copy of 2.17.x installed and see if the problem goes away.
>>6.       There’s always the possibility that Cacti is getting it wrong, so check there as well.
>>7.       Is it a bits/bytes issue?  If Cacti is showing 8x the MRTG value, maybe MRTG is showing in bytes. Though you have ‘options bits’.
>>8.       What frontend are you using for MRTG/RRD?  Some, notably Routers2.cgi in earlier versions, want the Target line to come first in the stanza.  Try rearranging them this way, maybe the Options line is getting ignored and you’re viewing in bytes/sec.
>> 
>>HTH
>> 
>>Steve
>> 
>>
>>________________________________
>> 
>>Steve Shipway
>>ITS Unix Services Design Lead
>>University of Auckland, New Zealand
>>Floor 1, 58 Symonds Street, Auckland
>>Phone: +64 (0)9 3737599 ext 86487
>>DDI: +64 (0)9 923 6487
>>Mobile: +64 (0)21 753 189
>>Email: s.shipway at auckland.ac.nz
>>PPlease consider the environment before printing this e-mail: 打印本邮件，将减少一棵树存活的机会
>> 
>>From:mrtg-bounces+s.shipway=auckland.ac.nz at lists.oetiker.ch [mailto:mrtg-bounces+s.shipway=auckland.ac.nz at lists.oetiker.ch] On Behalf Of Simon Westlake
>>Sent: Wednesday, 15 May 2013 2:26 p.m.
>>To: mrtg at lists.oetiker.ch
>>Subject: [mrtg] Strange MRTG problem
>> 
>>Hello,
>> 
>>It's been a long time since I used MRTG (I've been writing directly to rrds for a long time now) but an old employer of mine asked me to troubleshoot a legacy MRTG setup and I'm having a strange problem with it I cannot decipher.
>> 
>>I am graphing a port in two ways. One is in Cacti, one is with MRTG. Both are using 64bit counters (I have even setup the MRTG configuration to specifically use the HC OIDs rather than assuming they are being queried by specifying :::::2)
>> 
>>In Cacti, I get the correct value (approximately 700-800Mbps per second.) In MRTG, it is never more than 190-200Mbps.
>> 
>>My MRTG configuration looks like this:
>> 
>>WorkDir:/var/www/mrtg
>>Forks: 30
>>Refresh: 300
>>Interval: 5
>>MaxAge: 999999999999999999999999
>>SnmpOptions: retries => 1, timeout => 1
>>LogFormat: rrdtool
>> 
>> 
>>Options[DAS-c3229-bandwidth]: growright,bits,nolegend,nobanner,nopercent
>>Directory[DAS-c3229-bandwidth]: DAS-c3229
>>Target[DAS-c3229-bandwidth]: 1.3.6.1.2.1.31.1.1.1.10.140&1.3.6.1.2.1.31.1.1.1.6.140:8GF8961u at 24.160.238.254:::::2
>>Extension[DAS-c3229-bandwidth]: php
>>Suppress[DAS-c3229-bandwidth]: y
>>Title[DAS-c3229-bandwidth]: Bandwidth Utilization
>>Maxbytes[DAS-c3229-bandwidth]: 200000000000000000
>>AbsMax[DAS-c3229-bandwidth]: 200000000000000000
>> 
>>I have tried tweaking a lot of things and I can't seem to figure out what's going on. Any insight would be appreciated as I'm 2+ days into troubleshooting this thing and it's driving me crazy. Thanks!
>>-- 
>>- Simon Westlake
>
>
>
>-- 
>- Simon Westlake
>[ digitalgunfire.com ]
>AIM: ShirowDG
>ICQ: 2227213
>Voice: (414) 810-6392 
>_______________________________________________
>mrtg mailing list
>mrtg at lists.oetiker.ch
>https://lists.oetiker.ch/cgi-bin/listinfo/mrtg
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.oetiker.ch/pipermail/mrtg/attachments/20130515/961fdf8a/attachment.htm 