[j-nsp] PFE problems
Josef Buchsteiner
josefb at juniper.net
Fri Nov 26 11:17:07 EST 2004
Jonas,
this is jtree memory which is not the one which is complaining.
Please issue 'show chassis ssb' and look for the heap
utilization. It is the DRAM on the forwarding engine which is
reporting this message.
The "resolutions from iif 85 throttled" means you get too may
request from interface index 85 going out on Ethernet where we
have no next-hop/arp entry and we need to try to resolve this
and if we don't get an answer back and too many requests are
outstanding we will throttle this down. You can look at the
forwarding table which next-hop is in hold state then you know
where the traffic is supposed to get send and not getting an
answer back and you can look what is going on with the
arp-request/reply. You also know where the traffic is coming in
via index 85.
hope this helps
Josef
Wednesday, November 24, 2004, 10:26:29 AM, you wrote:
JF> Hi,
JF> i did that and the memory utilisation is fine:
JF> SSB0(ffm2-edge vty)# show jtree 0 memory
JF> Memory Statistics:
JF> 8388608 bytes total (4 banks)
JF> 2694504 bytes used
JF> 5694104 bytes free
JF> 8128 pages total
JF> 2611 pages used
JF> 5517 pages free
JF> 31 max freelist size
JF> Free Blocks:
JF> Size(b) Total(b) Free TFree Alloc
JF> -------- ---------- ---------- ---------- ----------
JF> 8 1689296 90 0 211072
JF> 16 541600 5 0 33845
JF> 24 1608 0 0 67
JF> 32 288 0 0 9
JF> 40 0 0 0 0
JF> 48 0 0 0 0
JF> 56 0 0 0 0
JF> 64 195456 1 1 3052
JF> 72 0 0 0 0
JF> 80 0 0 0 0
JF> 88 0 0 0 0
JF> 96 0 0 0 0
JF> 104 0 0 0 0
JF> Total 2428248
JF> Context: 0x3e0b14
JF> I am not sure about the blocks usage here however, since there seem no
JF> free ones?
JF> I am also getting some errors/warnings from the SSB itself:
JF> [Nov 24 09:24:37.412 LOG: Info] NH: resolutions from iif 85 throttled
JF> [Nov 24 09:25:11.577 LOG: Info] NH: resolutions from iif 85 throttled
JF> SSB0(ffm2-edge vty)# show jtree 0 sum
JF> Protocol Routes Bytes Used
JF> ------------- ---------- ----------
JF> IPv4 152262 2214536
JF> IPv6 659 11048
JF> MPLS 4 56
JF> Multi-service 1 16
JF> All in all it looks like there is enough memory left so that message
JF> shouldnt appear...
JF> Regards,
JF> Jonas
JF> On Wed, 2004-11-24 at 09:57, Piotr Marecki wrote:
>> Hi ,
>>
>> actually to see IPII lookup memory utilization one have to vty into FEB/SSB
>> and use "
>> one of the "cryptic" commands like "show jtree 0 memory" ,:
>> SSB0(lab-m20-1-re0 vty)# show jtree 0 memory
>> Memory Statistics:
>> 16777216 bytes total (4 banks)
>> 272480 bytes used
>> 16504736 bytes free
>> 16256 pages total
>> 264 pages used
>> 15992 pages free
>> 31 max freelist size
>>
>> Of course it would be better to have it all documented :)
>>
>> regards
>>
>> Piotr Marecki
>> Netia S.A
>>
>> ----- Original Message -----
>> From: "Bostjan Fele" <bostjan.fele at smart-com.si>
>> To: "Nisar Ali" <nisar at juniper.net>; "Jonas Frey" <jf at probe-networks.de>
>> Cc: <juniper-nsp at puck.nether.net>
>> Sent: Wednesday, November 24, 2004 9:42 AM
>> Subject: RE: [j-nsp] PFE problems
>>
>>
>> > That is memory on PFE, so on forwarding engine. Look for "show chassis
>> feb"
>> > and look for memory utilization.
>> >
>> > Regards,
>> > Bostjan
>> >
>> > --------------------------------------------------------
>> > Bostjan FELE
>> > CCIE #12594, InfoSec, JNCIS-M
>> > E-mail: bostjan.fele at smart-com.si
>> > Smart Com, d.o.o.
>> > Brnciceva 45, 1001 Ljubljana, Slovenia
>> > URL: http://www.smart-com.si
>> > Tel.: +386-1-5308-312
>> > Fax.: +386-1-5611-571
>> > Mob.: +386-41-399-363
>> > ---------------------------------------------------------
>> >
>> > -----Original Message-----
>> > From: juniper-nsp-bounces at puck.nether.net
>> > [mailto:juniper-nsp-bounces at puck.nether.net]On Behalf Of Nisar Ali
>> > Sent: Wednesday, November 24, 2004 5:13 AM
>> > To: Jonas Frey
>> > Cc: juniper-nsp at puck.nether.net
>> > Subject: Re: [j-nsp] PFE problems
>> >
>> >
>> > On Wed, Nov 24, 2004 at 01:25:25AM +0100, Jonas Frey wrote:
>> > > Hi,
>> > >
>> > > i am getting the following error all of a sudden on a M20:
>> > >
>> > > RT_PFE: NH IPC op 31 (CHANGE AGGREGATE NEXTHOP) failed, err 6 (No
>> > > Memory)
>> > >
>> > > What does this mean/refer to? The memory of the routing engine is hardly
>> > > 50% used.
>> >
>> > The message can be logged if, for any reason, the next hop change
>> > can not be completed (may be the next hop does not exist anymore). Any
>> > other logs before this error log started showing up? Any change
>> > in routes?
>> >
>> > What JUNOS you are running? Could you share configuration, "show
>> > chassis hard"? Do you have aggregated intefaces configured? If
>> > yes how many, with number of component links? Can you send "
>> > show route summary", "show route forwa summ".
>> >
>> > Thanks.
>> > nisar
>> >
>> > > Regards,
>> > > Jonas
>> > >
>> > > _______________________________________________
>> > > juniper-nsp mailing list juniper-nsp at puck.nether.net
>> > > http://puck.nether.net/mailman/listinfo/juniper-nsp
>> >
>> > --
>> > ================================================
>> > Nisar Ali
>> > Network Engineering Team (NET)
>> > ===============================================
>> > _______________________________________________
>> > juniper-nsp mailing list juniper-nsp at puck.nether.net
>> > http://puck.nether.net/mailman/listinfo/juniper-nsp
>> >
>> > _______________________________________________
>> > juniper-nsp mailing list juniper-nsp at puck.nether.net
>> > http://puck.nether.net/mailman/listinfo/juniper-nsp
>> >
>>
>> _______________________________________________
>> juniper-nsp mailing list juniper-nsp at puck.nether.net
>> http://puck.nether.net/mailman/listinfo/juniper-nsp
JF> _______________________________________________
JF> juniper-nsp mailing list juniper-nsp at puck.nether.net
JF> http://puck.nether.net/mailman/listinfo/juniper-nsp
More information about the juniper-nsp
mailing list