[j-nsp] JUNOS 6.0R1.3 traffic statistic is not accurate?

[bbird at epik.net]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

We are seeing similar behavior with Junos6.0R1.5 on the M160 platform wrt
POS OC192 links.  In fact, our SNMP poller is reporting several instances
per day of insane delta values on counters, and discarding the polls.  We've
upgraded 2 out 10 M160 that previously ran Junos 5.3R2.4, and have made no
changes to the poller configuration.  The remaining 8 M160's still running
5.3R2.4 are not exhibiting these errors.  The example below is
ifHCInUcastPkts, however this also seems to be affecting ifHCInOctet and
ifHCOutOctets, as well.  Our latest weekly utilization report is showing
823.53% utilization now.

Monday, 10/13/2003 16:37:41
 Received large delta from 'jcvlflnj2m6-LW-so-3/0/0'. Poll is dropped (OID
in error is ifHCInUcastPkts. Delta is 1.844674406e+19. Old value is
2.288484296e+10. (Counter64: 0x5 0x540afdd1) Current value is
1.718154249e+10. (Counter64: 0x4 0x198858)).


- -Ben 

#-----Original Message-----
#From: Yu-lin Chang [mailto:ylchang at ascc.net] 
#Sent: Thursday, October 16, 2003 10:16 PM
#To: juniper-nsp at puck.nether.net
#Subject: [j-nsp] JUNOS 6.0R1.3 traffic statistic is not accurate?
#
#
#Hi,
#
#After upgrade to JUNOS 6.0R1.3 on my M160, I got something
#wrong on the traffic statistics. The traffic statistics
#are calculated by MRTG through SNMP (use 64bit counter). 
#I did not modify MRTG config after upgrade the JUNOS.
#
#I divide an ATM-OC12 into 5 PVCs (without traffic shaping).
#The traffic statistics on the interface is not equal to 
#the sum of this 5 PVCs. For example, the traffic on ATM-OC12 
#interface is 100mbps, but the sum of 5 PVCs goes to 2 or 3 times
#higher (200mbps+).
#
#The same problem appears on the STM-1 and STM-4 SONET interfaces.
#I also divide SONET interface into 3 pipes by using different 
#DLCI numbers on Frame-Relay Encapsulation.
#
#The traffic statistics on STM-1 or STM-4 interface have the same
#behavior like the ATM-OC12 interface. The sum of traffics of
#the DLCIs is much higher then the traffic on interface.
#
#Anyone meets the same problem? Does this a bug of JUNOS 6.0R1.3? 
#Thanks in advance.
#
#BR,
#Yu-lin
#--
#
#P.S. my ATM and SONET interfaces hardware info are listd bellow.
#
#> show chassis pic fpc-slot 0 pic-slot 0    
#PIC fpc slot 0 pic slot 0 information:
#  Type                             1x OC-12 ATM, MM
#  ASIC type                        K chip
#  State                            Online    
#  PIC version                  1.5
#  Uptime                         19 days, 6 hours, 46 minutes, 1 second
#
#> show chassis pic fpc-slot 1 pic-slot 0    
#PIC fpc slot 1 pic slot 0 information:
#  Type                             4x STM-1 SDH, SMIR
#  ASIC type                        D chip
#  State                            Online    
#  PIC version                  2.4
#  Uptime                         19 days, 6 hours, 46 minutes, 
#5 seconds
#
#> show chassis pic fpc-slot 2 pic-slot 0    
#PIC fpc slot 2 pic slot 0 information:
#  Type                             4x OC-12 SONET, SMIR
#  ASIC type                        D chip
#  State                            Online    
#  PIC version                 1.13
#  Uptime                         19 days, 6 hours, 46 minutes, 
#5 seconds
#-----
#
#_______________________________________________
#juniper-nsp mailing list juniper-nsp at puck.nether.net
#http://puck.nether.net/mailman/listinfo/juniper-nsp
#

-----BEGIN PGP SIGNATURE-----
Version: PGP 8.0

iQA/AwUBP49uhNFQh6ARB7TZEQJBwgCfUoFYlisfsnuC1sesKaqokC3LFBcAnA7H
TrUVRlJ97o8+qSNRIR7dQSbb
=/uGO
-----END PGP SIGNATURE-----