RE: Re: Initial release of the Bacula SNMP Subagent - v1.2

On Mon, 2005-10-24 at 03:10, Dave Shield wrote:
> On Fri, 2005-10-21 at 17:32 -0500, Les Mikesell wrote:
> > What's supposed to happen when the value doesn't fit in a Counter32?
> 
> The Counter32 value will report the bottom 32-bits
> (with the higher bits simply being discarded).

Thanks - it isn't doing that.

> Remember that in the SNMP world, an individual Counter-based value
> has no meaning.  Counter32 (and Counter64) object are used to
> monitor the *difference* between two successive readings.
> 
> 
> So:
>     ifconfig eth1 ; sleep 60 ; ifconfig eth1
> 
> might report
>     eth1   TX bytes:33033390619 (31503.0 Mb)
>     eth1   TX bytes:33033391619 (31503.0 Mb)
> 
> 
> while
>     snmpget eth0 ; sleep 60 ; snmpget eth0
> 
> might report
> 
>     ifOutOctets.1 = 149182455
>     ifOutOctets.1 = 149183455
> 
> 
> The raw numbers are different, but the delta:
> 
>   33033391619 - 33033390619  = 1000
>   149183455   - 149182455    = 1000
> 
>  is the same.   That's how SNMP Counters work.

The numbers I posted before were real, and still doing the same
thing:
ifconfig says:  TX bytes:34697827021 (33090.4 Mb)    (changed)
snmpwalk says:  IF-MIB::ifOutOctets.3 = Counter32: 0 (still zero)


This is on Centos3.x (should match RHEL3 with updates) in package
net-snmp-5.0.9-2.30E.19.  I suppose I can try building a newer version.

-- 
  Les Mikesell
    [email protected]




-------------------------------------------------------
This SF.Net email is sponsored by the JBoss Inc.
Get Certified Today * Register for a JBoss Training Course
Free Certification Exam for All Training Attendees Through End of 2005
Visit http://www.jboss.com/services/certification for more information