On 2019/08/13 13:43, Martijn van Duren wrote:
> On 8/13/19 1:40 PM, Claudio Jeker wrote:
> > On Tue, Aug 13, 2019 at 01:27:44PM +0200, Martijn van Duren wrote:
> >>
> >>
> >> On 8/13/19 1:23 PM, Stuart Henderson wrote:
> >>> On 2019/08/13 12:20, Stuart Henderson wrote:
> >>>> On 2019/08/13 11:30, Martijn van Duren wrote:
> >>>>> Found this one by because snmpd returns an invalid varbind on my laptop.
> >>>>> Doing some digging I found the following:
> >>>>> Error message:
> >>>>> mib_iftable: iwm0: invalid ifq: Operation not supported
> >>>>> mib_iftable: em0: invalid ifq: Operation not supported
> >>>>>
> >>>>> sysctl mib:
> >>>>> { CTL_NET, PF_INET, IPPROTO_IP, IPCTL_IFQUEUE, IFQCTL_DROPS };
> >>>>>
> >>>>> file: usr.sbin/snmpd/mib.c:1198
> >>>>>
> >>>>> I'll look at why snmpd returns an invalid packet later, but did anything
> >>>>> change in packet drop land that could've caused this?
> >>>>>
> >>>>> martijn@
> >>>>>
> >>>>
> >>>> Yes, the input queuing mechanism was changed, these are no longer used.
> >>>>
> >>>
> >>> oops s/input/interface :)
> >>>
> >> Does this mean we need to remove this mib from snmpd, or do we need to
> >> invoke some other incantation?
> >
> > Uhm wait, that is querying net.inet.ip.ifq.drops and yes, that is gone but
> > that has nothing todo with interface counters. What kind of mib value is
> > that?
> >
>
> That would be:
> .iso.org.dod.internet.mgmt.mib_2.interfaces.ifTable.ifEntry.ifInDiscards
Prior to this sysctl mib removal, the snmp mib was already returning bogus
information here.
It should be this:
"The number of inbound packets which were chosen to be discarded even
though no errors had been detected to prevent their being deliverable
to a higher-layer protocol. One possible reason for discarding such a
packet could be to free up buffer space."
I think this should probably display the same as Idrop in
"netstat -I $interface -nd".
