And finally the test procdure (and initial results):

[Peter Van Epp]

	Yep thats about what I figured. We are by and large in good shape and
should (or at least could, I used the single name to print two field ones out
of laziness :-)) split everything in to src and dst. Once we get test cases
identical we should be able to dump the output from 2.0.6 and 3.0 to files 
and feed them to a perl script that can scan for mismatches to catch any 
corner cases that are about (and find folks with vlan, mpls and user data none
of which I have any large volumes of). I was thinking about boosting that buffer
to 4096 but 1024 was enough for 42 labels :-) and I expect thats where your 
seg fault was. By the way "status" in the ra3.conf.full file should be "state"
to print the correct field. I caught it in the man page but not in the conf
file :-).

Peter Van Epp / Operations and Technical Support 
Simon Fraser University, Burnaby, B.C. Canada


On Mon, Jun 19, 2006 at 05:25:46PM -0400, Carter Bullard wrote:
> Hey Peter,
>    What a great test!!!!!    argus-3.0 has 132 printable fields, and  
> so I'm applying
> your approach to test all of them at once, which is now causing me to  
> have to
> fix yet another segmentation fault!!!!
> 
> So, we have some field name changes,  I'd rather not put back the old
> names, because in some cases we don't support the old function.  I've
> removed fields that print 2 fields, as we now sort and merge on many of
> these fields, and having to specify either the source or the destination
> is important,  so I split them.   Some of the multiple field keywords  
> are
> still there, but I should remove them for completeness.  As as example,
> 'win' is now 'swin' and 'dwin', and win by itself doesn't work, but  
> 'mac'
> by itself is still there, (definately a mistake).  I'll see if I can  
> generate
> the list here:
> 
>    Single field converted to 's' and 'd':
>       win
>       pkts
>       bytes
>       load
>       loss
>       rate
>       mac
>       mpls
>       vlan
>       ipid
>       user
> 
>    Field name change:
>       ind       -   flgs    print flags
>       status  -   state  print status
> 
> This seems like a good list.
> 
>    OK, so we're not picking up the IP attributes from the 2.0  
> records, ttl, tos, ipid,
> and options.  we have the bps values, not sure why we're not printing  
> them out,
> probably looking for a specific DSR structure and the pointer may not be
> there.   The rest looks good, because we need to correct for the name  
> changes?
> 
> Carter
> 
> 
> 
> On Jun 19, 2006, at 2:20 PM, Peter Van Epp wrote:
> 
> >ra -F ra2.conf.full -r t.argus
> >StartTime,LastTime,Trans,Dur,AvgDur,SrcAddr,DstAddr,Type,Sport,Dport,S 
> >rcTOS,DstTOS,SrcTTL,DstTTL,SrcBytes,DstBytes,SrcPkt,DstPkt,Src_bps,Dst 
> >_bps,Src_Loss,Dst_Loss,Src_pps,Dst_pps,ProbeId,Flgs,SrcMacAddr,DstMacA 
> >ddr,Dir,SrcJitter,DstJitter,State,srcUdata,dstUdata,SrcWin,DstWin,Seq, 
> >sMPLS,dMPLS,sVLAN,dVLAN,IpId
> >1132974665.005254,1132974683.159241,,18.153987,18.153987,229.97.122.20 
> >3,1,man,v2.0,0,0,0,0,0,0,0,0,0,-0.00,-0.00,,0.00,0.00,3848370891,,,,,0 
> >.000,0.000,STA,,,,,0,,,,,
> >1149490800.390902,1149490800.392313,1,0.001411,0.001411,206.127.21.254 
> >,142.58.207.207,udp, 
> >65442,123,0,16,46,61,90,90,1,1,510276.40,510276.40,0.0000,0.0000,0.00, 
> >0.00,3848370891,,0:90:69:c0:e0:1f,0:e0:63:13:7e:0,<->, 
> >0.000,0.000,CON,,,,,1307546097,,,,,0xbc5f
> >
> >
> >%ra3 -F ra3.conf.full -n -r t.argus
> >StartTime,LastTime,Trans,Dur,AvgDur,SrcAddr,DstAddr,Proto,Sport,Dport, 
> >sTos,dTos,sTtl,dTtl,SrcBytes,DstBytes,SrcPkts,DstPkts,Src_pps,Dst_pps, 
> >SrcLoss,DstLoss,Src_bps,Dst_bps,SrcId,SrcMac,DstMac,Dir,SrcJitter,DstJ 
> >itter,srcUdata,dstUdata,SrcWin,DstWin,Seq,sMpls[0],dMpls 
> >[0],sVlan,dVlan,sIpId,dIpId
> >1149490800.390902,1149490800.392313,1,0.001411,0.001411,206.127.21.254 
> >,142.58.207.207,udp,65442,123, , , , , 
> >90,90,1,1,708.717,708.717,0,0,510276.375,510276.375,229.97.122.203,0:9 
> >0:69:c0:e0:1f,0:e0:63:13:7e:0,<->,,,,, , ,1307546097 ,  , ,
> >
> >