[j-nsp] M160/JunOS 7.6R1.10: ae0 fails to install L2 descriptor
peter brenner
peter5192 at hotmail.com
Thu Dec 28 22:49:54 EST 2006
Hi,
tonight I was trying to add a third gigE link to a 2gbit/s aggregated gigE
connection when the following happened:
Dec 29 02:53:55 ham-cr2-re1 chassisd[2968]: CHASSISD_SNMP_TRAP10: SNMP trap
generated: FRU power on (jnxFruContentsIndex 8, jnxFruL1Index 2,
jnxFruL2Index 4, jnxFruL3Index 0, jnxFruName PIC: @ 1/3/*, jnxFruType 11,
jnxFruSlot 2, jnxFruOfflineReason 2, jnxFruLastPowerOff 517772085,
jnxFruLastPowerOn 518744346)
Dec 29 02:54:05 ham-cr2-re1 chassisd[2968]: CHASSISD_IFDEV_CREATE_NOTICE:
create_pics: created interface device for ge-1/3/0
Dec 29 02:54:05 ham-cr2-re1 /kernel: ae_link_op: link ge-1/3/0.2 (lidx=2)
attached to bundle ae0.2
Dec 29 02:54:05 ham-cr2-re1 /kernel: ge-1/3/0.2 joins ae0.2
...
Dec 29 02:54:06 ham-cr2-re1 /kernel: ae_link_op: link ge-1/3/0.75 (lidx=2)
attached to bundle ae0.75
Dec 29 02:54:06 ham-cr2-re1 /kernel: ge-1/3/0.75 joins ae0.75
Dec 29 02:54:20 ham-cr2-re1 /kernel: RT_PFE: NH IPC op 2 (CHANGE NEXTHOP)
failed, err 1 (Unknown)
Dec 29 02:54:20 ham-cr2-re1 /kernel: RT_PFE: NH IPC op 1 (ADD NEXTHOP)
failed, err 1 (Unknown)
Dec 29 02:54:20 ham-cr2-re1 /kernel: RT_PFE: NH details: idx 20985 type 2
ifl 142
Dec 29 02:54:20 ham-cr2-re1 /kernel: RT_PFE: NH IPC op 31 (CHANGE AGGREGATE
NEXTHOP) failed, err 5 (Invalid)
Dec 29 02:54:20 ham-cr2-re1 /kernel: RT_PFE: NH IPC op 1 (ADD NEXTHOP)
failed, err 1 (Unknown)
Dec 29 02:54:20 ham-cr2-re1 /kernel: RT_PFE: NH details: idx 20986 type 2
ifl 142
Dec 29 02:54:20 ham-cr2-re1 /kernel: RT_PFE: NH IPC op 31 (CHANGE AGGREGATE
NEXTHOP) failed, err 5 (Invalid)
Dec 29 02:54:20 ham-cr2-re1 /kernel: RT_PFE: NH IPC op 1 (ADD NEXTHOP)
failed, err 1 (Unknown)
Dec 29 02:54:20 ham-cr2-re1 /kernel: RT_PFE: NH details: idx 20987 type 2
ifl 142
...
Dec 29 02:54:20 ham-cr2-re1 sfm3 NH(nh_ucast_add): Fails to install L2
descriptor
Dec 29 02:54:21 ham-cr2-re1 sfm3 NH(nh_ucast_add): Fails to install L2
descriptor
Dec 29 02:54:21 ham-cr2-re1 sfm3 NH: Aggregate nh (7927) with unknown nh in
list (20985)
Dec 29 02:54:21 ham-cr2-re1 sfm3 NH: Non-existant NH (20985:Unicast, 2) in
generic change path
...
and then the other sfms quickly chimed in:
Dec 29 02:54:44 ham-cr2-re1 sfm0 NH: Aggregate nh (4458) with unknown nh in
list (22962)
Dec 29 02:54:44 ham-cr2-re1 sfm2 NH: Aggregate nh (10081) with unknown nh
in list (23518)
Dec 29 02:54:45 ham-cr2-re1 sfm1 NH: Aggregate nh (12745) with unknown nh
in list (24054)
Dec 29 02:54:45 ham-cr2-re1 sfm0 NH: Aggregate nh (8768) with unknown nh in
list (23343)
Dec 29 02:54:45 ham-cr2-re1 /kernel: RT_PFE: NH IPC op 1 (ADD NEXTHOP)
failed, err 1 (Unknown)
Dec 29 02:54:45 ham-cr2-re1 /kernel: RT_PFE: NH details: idx 26128 type 2
ifl 201
Dec 29 02:54:45 ham-cr2-re1 /kernel: RT_PFE: NH IPC op 1 (ADD NEXTHOP)
failed, err 1 (Unknown)
Dec 29 02:54:45 ham-cr2-re1 /kernel: RT_PFE: NH details: idx 26129 type 2
ifl 209
Dec 29 02:54:45 ham-cr2-re1 /kernel: RT_PFE: NH IPC op 29 (ADD AGGREGATE
NEXTHOP) failed, err 5 (Invalid)
Dec 29 02:54:45 ham-cr2-re1 sfm0 NH: Aggregate nh (9121) with unknown nh in
list (23378)
Dec 29 02:54:45 ham-cr2-re1 sfm1 NH: Aggregate nh (13026) with unknown nh
in list (24156)
Dec 29 02:54:45 ham-cr2-re1 sfm1 NH: Aggregate nh (13074) with unknown nh
in list (24171)
Dec 29 02:54:45 ham-cr2-re1 sfm2 NH: Aggregate nh (12588) with unknown nh
in list (24014)
Dec 29 02:54:45 ham-cr2-re1 sfm0 NH: Aggregate nh (9459) with unknown nh in
list (23419)
Dec 29 02:54:45 ham-cr2-re1 sfm3 NH: Aggregate nh (14014) with unknown nh
in list (25161)
Dec 29 02:54:46 ham-cr2-re1 sfm3 NH: Aggregate nh (15076) with unknown nh
in list (25184)
Dec 29 02:54:46 ham-cr2-re1 sfm0 NH: Aggregate nh (9669) with unknown nh in
list (23454)
Dec 29 02:54:46 ham-cr2-re1 sfm3 NH(nh_ucast_add): Fails to install L2
descriptor
de-pb at ham-cr2-re1> show log messages.1.gz | match "unknown nh in list" |
count
Count: 604 lines
de-pb at ham-cr2-re1> show log messages.1.gz | match "NH IPC op" | count
Count: 22527 lines
de-pb at ham-cr2-re1> show log messages.1.gz | match "Fails to install L2" |
count
Count: 816 lines
Looking at the direct vty output it looks like one of the nhdb zones for the
L2 Descriptors was full.
[hoping that hotmail won't butcher the table]
Chip Start Size Rsvd Used/Hi Water/Total Size Name
---- ----- ----- ----- -------------------- ---- ----
0 20000 01000 00000 0/0/512 8 Multicast RTP
0 21000 02400 00000 0/0/9216 1 Multicast Lists
0 23400 04400 00006 176/177/17408 1 Next-Hop Entries
0 27800 18800 0004b 926/930/50176 2 L2 Descriptors
0 40400 001ff 00000 8/8/63 8 L2 Programs
1 20000 01000 00000 0/0/512 8 Multicast RTP
1 21000 02400 00000 0/0/9216 1 Multicast Lists
1 23400 04400 00006 8174/12468/17408 1 Next-Hop Entries
1 27800 18800 0004b 32674/50061/50176 2 L2 Descriptors
1 40400 001ff 00000 10/10/63 8 L2 Programs
6 20000 01000 00000 0/0/512 8 Multicast RTP
6 21000 02400 00000 0/0/9216 1 Multicast Lists
6 23400 04400 00006 0/0/17408 1 Next-Hop Entries
6 27800 18800 0004b 0/0/50176 2 L2 Descriptors
6 40400 001ff 00000 1/1/63 8 L2 Programs
I have a couple of questions:
a) Is it normal to have 32k L2 Descriptors for 8.2k Next-Hop Entries?
b) Is there a way to increase the number of available L2 Descriptors? (how
many L2 Descriptors does a SFM-16 support?)
c) Is there a way to make the router fail with less impact to the network
(for example simply shutting down the new interface automatically instead of
refusing to update the next-hop table until the interface is taken down and
all sfms are restarted manually)
Any help is appreciated.
Thanks in advance, Peter
_________________________________________________________________
FREE pop-up blocking with the new MSN Toolbar - get it now!
http://toolbar.msn.click-url.com/go/onm00200415ave/direct/01/
More information about the juniper-nsp
mailing list