[j-nsp] VCP-ports with missing neighbors on QFX

Franz Georg Koehler lists at openunix.de
Fri Jan 13 12:55:38 EST 2017


After some research, this seems to be a bug (BTW, this is on: 14.1X53-D30.3)
We see a lot of those messages:

fpc4 dcbcm_check_stuck_buffers: Buffers are stuck on queue 0 of port 69 
fpc4 Soft reset port 69 due to stuck buf 
fpc4 MacDrainTimeOut:port 0,hg1, timeout draining packets (131068 cells remain) 
fpc4 : dcbcm_drv_port_get_link_status : MmuDrain fail for port 69 : reason Internal error 
fpc4 MacDrainTimeOut:port 0,hg1, timeout draining packets (131068 cells remain) 
fpc4 : dcbcm_drv_port_get_link_status : Port disable fail for port 69 : reason Internal error 
fpc3 dcbcm_check_stuck_buffers: Buffers are stuck on queue 0 of port 69 
fpc3 Soft reset port 69 due to stuck buf 
fpc3 MacDrainTimeOut:port 0,hg1, timeout draining packets (131064 cells remain) 
fpc3 : dcbcm_drv_port_get_link_status : MmuDrain fail for port 69 : reason Internal error 
fpc3 MacDrainTimeOut:port 0,hg1, timeout draining packets (131064 cells remain) 
fpc3 : dcbcm_drv_port_get_link_status : Port disable fail for port 69 : reason Internal error 
fpc6 dcbcm_check_stuck_buffers: Buffers are stuck on queue 7 of port 69 
fpc6 Soft reset port 69 due to stuck buf 
fpc6 MacDrainTimeOut:port 0,hg1, timeout draining packets (131070 cells remain) 
fpc6 : dcbcm_drv_port_get_link_status : MmuDrain fail for port 69 : reason Internal error 
fpc6 MacDrainTimeOut:port 0,hg1, timeout draining packets (131070 cells remain) 

Port 69/hg1 seems to be the VCP-Port that is not operational any more:
TFXPC3(vty)# set dc bc "phy info"

HW (unit 0)
Phy mapping dump:
      port   id0   id1  addr iaddr                    name    timeout
(...)
 xe47( 48)  600d  84f9    4e   197                BCM84848     250000  
 xe48( 49)  600d  8770    e9    e9             TSC-A2/18/4     250000  
 xe49( 50)     0     0    ff    ea                    Null     250000  
 xe50( 51)     0     0    ff    eb                    Null     250000  
 xe51( 52)     0     0    ff    ec                    Null     250000  
 xe52( 53)  600d  8770    e5    e5             TSC-A2/17/4     250000  
 xe53( 54)     0     0    ff    e6                    Null     250000  
 xe54( 55)     0     0    ff    e7                    Null     250000  
 xe55( 56)     0     0    ff    e8                    Null     250000  
 xe56( 57)  600d  8770    ed    ed             TSC-A2/19/4     250000  
 xe57( 58)     0     0    ff    ee                    Null     250000  
 xe58( 59)     0     0    ff    ef                    Null     250000  
 xe59( 60)     0     0    ff    f0                    Null     250000  
 xe60( 61)  600d  8770    f1    f1             TSC-A2/20/4     250000  
 xe61( 62)     0     0    ff    f2                    Null     250000  
 xe62( 63)     0     0    ff    f3                    Null     250000  
 xe63( 64)     0     0    ff    f4                    Null     250000  
  hg0( 65)  600d  8500    70   1a1                BCM84328     250000  
 xe64( 66)     0     0    71   1a2                    Null     250000  
 xe65( 67)     0     0    72   1a3                    Null     250000  
 xe66( 68)     0     0    73   1a4                    Null     250000  
  hg1( 69)  600d  8500    74    8d                BCM84328     250000  
 xe67( 70)     0     0    75    8e                    Null     250000  
 xe68( 71)     0     0    76    8f                    Null     250000  
 xe69( 72)     0     0    77    90                    Null     250000  




Does anybody have an idea besides complete reboot (and upgrading the software at that chance)?



Best regards,

Franz Georg Köhler


More information about the juniper-nsp mailing list