One possible workaround in this scenario might be (obviously depends on
IGP complexity etc):
Run OSPF temporarily, carry v6 routes in there, then you can reconfigure
IS-IS as you see fit. Once you're done, disable OSPF again.
Regards,
Felix
On 26.06.12 15:18, Jared Gull wrote:
After digging
On the other hand - with a sufficiently low sampling rate, RE based
sampling may be feasible. No license needed for that.
We use RE based sampling on MX. Works for us, YMMV.
We used RE-based sampling (with 1:1000 sampling), which worked just fine
since the M5 days. Sadly, there is no
Ryan,
On 18.07.11 02:59, Ryan Finnesey wrote:
We do have a Juniper account rep so I will have a chat with her. I would
much prefer to deal directly with JTAC. It has been about 6 years since I
dealt with Dell but I did not find their support all that great.
as Dell switches their OEM
Paul,
by default JUNOS will do per-prefix balancing with multipath. Did you
configure load-balance per-packet (which actually is per flow and
named this for historical reasons) on forwarding table export?
For the load balancing to put multiple paths to the same target into the
forwarding
On 21.12.10 17:20, ibariouen khalid wrote:
Dear all
Can someone tell me what is the default value of default_arp_policer ? is
there any recommendation to reduce the values ;because i have an issue with
an ARP storme and the router is impacted .
the major problem with the default arp
Charlie,
We only have 165 terms.
applied in which manner to which ports? Depending on how you do this
there could be a multiplication issue, we ran into issues with that
(albeit in a 9.4 release IIRC) with even fewer filters.
regards,
Felix
--
Felix Schüren
Head of Network
CB,
Aug 25 12:13:36 ALBQ_EX4500 mib2d[861]: SNMP_TRAP_LINK_DOWN: ifIndex 536,
ifAdminStatus up(1), ifOperStatus down(2), ifName xe-0/0/16
Aug 25 12:14:04 ALBQ_EX4500 last message repeated 20 times
Aug 25 12:16:06 ALBQ_EX4500 last message repeated 61 times
Aug 25 12:26:05 ALBQ_EX4500 last
On 01.10.10 07:26, Michel de Nostredame wrote:
Hi,
I was checking my EX4200 trying to resolve a strange connection
problem with my vendor through a Metro Ethernet.
During that time I found another weird situation (it is not related to
the metroEthernet connection).
I setup two topology
Henri Khou wrote:
Hello,
I'm trying to find the best way to redistribute BGP default route to 2
independant organizations running OSPF with a different cost for each
organization.
I am used to Cisco IOS where you can run 2 OSPF instance by specifying a
process ID for every OSPF instance.
Chris,
#1 - I have two eBGP neighbors using BFD. One of the neighbors tripped, now
BFD won't re-establish. BGP is up however.
#2 - I'm using IRB interfaces on the MX platform. After the failover,
traffic will not forward.. You can communicate RE to host, but HOST to HOST
on the same box
Charlie,
what L2 load balancing hashing is done on the EX 4200 series?
There is no forwarding-options hash-key configuration statement.
I have two interfaces, ae0 and ae1 that run BGP in a 2 member VC stack.
Behind the VC stack is a row of PCs with their own LAG aeX's.
Will ae0 and
Gabriel,
I need to do traffic engineering in backbone IP/MPLS from point A to point
B, but just need some prefixes, part of the traffic is sent via TE and I'm
seeing all traffic exiting via TE from point A to B. *What do I need to send
only the prefix 10.152.42.0/25 via TE and not all traffic*?
Luis,
Hello ..
I´m working with isis using iso addressing, so now when i see the routes in
my EX , it has the next:
juni...@junex.cvie.mgmt.01# run show route
inet.0: 13 destinations, 13 routes (13 active, 0 holddown, 0 hidden)
+ = Active Route, - = Last Active, * = Both
Eric,
As for the L1 route preference, that's what I don't understand. If
R1/R2 are getting each other's loopbacks through L2 with a preference
of 18, but then I swap the L1/L2 preferences so that L2 now has a
pref of 15, why would the L1 route always get preferred?
IS-IS always prefers
Bill Blackford wrote:
Platform: EX3200-24T
JUNOS: 10.0S1.1
PROBLEM:
I have several EX series devices deployed and so far, and only one is
exhibiting the behavior of the routing-engine uptime being vastly different
from the actual system uptime.
When I issue:
show system uptime
I'm seeing
Paul,
When I do show route community 11666:5000 I get a list of the proper
prefixes as expected. 11666:5000 is our own network routes. The same
occurs when I list 11666:4000 which is customer network routes - displays
the list correctly. This confirms in my mind that the MX480 is receiving
Kessler, Ben wrote:
Hello -
I'm in the process of upgrading a lot of J6350 routers and have one that is
failing on boot after the upgrade. I'm getting the following messages on
the console when the router tries to boot:
OK reboot
Rebooting...
elf32_loadfile: can't load module before kernel
Taqdir Singh wrote:
Hi all,
I have read that in juniper vrrp has advertise time of 1 sec and
holdtime of 0 sec by default ?
generally, VRRP waits for (3 * advertise time) (as per the RFC). I never
checked if the Skew_Time (lower priority makes you wait longer) is
implemented in
Taqdir Singh wrote:
hi Felix,
here is the link where its mentioned hold time by default is 0
Ah, that's hold-time for preemption. When a former vrrp master router
rejoins the network after a reboot, for example, if preemption is
enabled (default) then it immediately takes over, regardless of
Tore Anderson wrote:
* Richard A Steenbergen
Correct. I actually found some old gripes about this when I searched
j-nsp after noticing the problem, but it is a big enough issue that I
think it needs to be repeated again (and again and again, until it
gets fixed :P).
I'll be happy to join the
Nonsense. The CPU usage might be ever so slightly higher in v9, but
neither is a major contributor to overall load even at the maximum
supported RE sampling rate. The only time you *need* a services card is
when you want to do a much higher rate than the RE can support (i.e. you
want to do some
Uttam,
On juniper M10i with JUNOS 9.2, we have flow exported by the routing engine
sampling packets headers and had aggregated them into flows.We have two
upstream and peered number of customers, we have packet sampling done by
defining a firewall filter to accept and sample all traffic and
Ross,
On Tue, Feb 02, 2010 at 11:56:56AM +0100, Felix Sch?ueren wrote:
show arp no-resolve | match x/y/z | save tempfile
(which takes ~5 seconds on an ethernet router with just ~20k ARP entries)
start shell
awk '{print clear arp hostname $2}' tempfile
and then copy-paste hundreds or
To reply to myself:
start shell
cli -c 'show arp no-resolve' | grep 'x/y/z' | awk '{ print clear arp
hostname $2 }' | cli
start shell
arp -an | grep 'x/y/z' | awk...
the arp information is available in the shell directly :)
-f
--
Felix Schüren
Head of Network
Sven,
firewall {
##
## Warning: configuration block ignored: unsupported platform
(ex4200-24f)
##
filter REF {
term snmp {
from {
Does this generally not work on the EX-series
or just not for the -24f?
edit firewall family inet filter
On 27.01.10 08:31, Nils Kolstein wrote:
The E-track is the Enterprise routing track and focusses on J-series
routers. So some specific J-series hardware-related issues are part
of this track whereas the M-track focuses on the M/T-series.
you're referring to the -ER (enterprise routing) track.
David,
On 07.01.10 18:31, David Ball wrote:
GRES with NSR, yes. Apparently the stalling has to do with the
master RE not receiving 'ok' from backup RE when it says it has an
update. It won't install the new route to the forwarding table until
the 'ok' is received from the backup, or
David,
I'm working with ATAC tonight to get them a running kernel core dump
so they can look for root cause, but apparently disabling GRES,
committing, re-enabling GRES, and committing again, somehow can
temporarily resolve the issue (get the routes installed, I guess
?!?!). Don't ask me how
Eric,
Eric Van Tol wrote:
Hello,
I'm trying to bring up ISIS between a J2350 running 8.5R2.10 and an MRV
Optiswitch. I currently have ISIS running on the J2350 in an IPv6-only
capacity. If I enable it for this new interface and disable IPv6 in the
interface config in ISIS, I still see
Brendan Mannella wrote:
Upgrading to 9.3R4.4 seems to have broken my rate-limiting.
The switch is being used to aggregate colo customers, and i need to be able
to create rate-limits for different speeds and apply them to different
physical ports.
Can anyone provide a example config
Dan,
firewall {
policer 500m {
if-exceeding {
bandwidth-limit 600m;
burst-size-limit 15m;
}
then discard;
}
family inet {
filter 500m-limit {
term default {
then policer 500m;
Ramesh Karki wrote:
Hello all,
We are facing some issues on Juniper M10i, i.e.
First, we had to hard reset the bgp peer whenever we change the policies
(inbound policy) that we had set. By just doing soft reset the router will
not take effect of that changed policies until we do hard
Just reading the release notes for 10.0 I found myself thinking Yes!
finally! An interface range command for the EXes!, only to learn that
this is a strange mixture of a concept somewhere between apply-group
interface-set - it's NOT the simply yet incredibly effective cisco-style
which would
Hoogen,
Okay.. Earlier task required while accepting routes from peer to tag
them with a community and prepend them with as number 65412 twice...
I notice that when I deactivate that.. It works.. So obviously R3 is
considering the routes received from R1 with prepend of 65412
Hoogen wrote:
Hi Felix,
Thank you for the reply..
I am not sure how that 17 hidden routes came into play... But its not
there now.. I still see the issue..
hm. Do you have an as-loop? routes with as-loops don't show up in the
rib-in. Please paste r1's protocols bgp stanza, and
Quoc Hoang wrote:
What cabinets do you have the switch installed in? The 19' or 24'
cabinets? Since it's side to side cooling, I'm afraid the cold hot
aisle layout in a typical datacenter may pose an issue.
600x1200mm racks, hot/cold aisle, yes, it's problematic. Totally
braindead decision by
Quoc,
Hi,
I'm considering purchasing the big iron EX8216 chassis for our core
switching. These switches are pretty brand new to the market. Anyone
running these and care to share their operational
experience/feedback/issues seen so far?
I can only comment on the EX8208 - running okayish so
and also note on the 3200's redundant power supplies is available -
but the 2nd PS is external
I suppose it's best to just state that redundant power is not available
for the 3200 - in my experience, the whole external power supply stuff
is causing more problems than it's solving.
around
Paul,
Does anyone have any real-world feedback on the layer3 performance of EX3200
and/or EX4200 switches? I've searched around and cannot find out the
capabilities (pps/Gbps) but new to the Juniper world ;)
we've been running a couple of 2-member virtual chassis in production
for some
All,
just a quick heads up: flowfilter (aka flowspec aka inetflow) does not
work with ipv6. And it took JTAC just two weeks to figure it out...
JTAC wrote:
[...] flow filters in IPv6 are not supported
A shame, really, as flowfilter is one of the best features ever. I hope
they'll add v6
Brendan,
When you say disabled LACP does that mean make both sides passive?
Or one side active and the other side passive?
I tryed searching the docs, but all i got out of it was that if both
sides are set to passive the link will not automatically come up. What
exactly does that mean? And
Jeff S Wheeler wrote:
I posted about this back in September of 2008, but I don't think anyone
took interest at that time, perhaps because I was looking at the issue
on nothing more recent than JUNOS 8.5R1.14. I noticed today this is
still happening in JUNOS as recent as 9.3R2.8. I bet if I
It'll try to use it, but it shouldn't work for actual forwarding - you
can't get packets from the PFE towards fxp0 (but be warned: you can get
packets from fxp0 to the PFE...).
Kind regards,
Felix
Patrik Olsson wrote:
Hi,
sorry for typing faster than thinking :-)
If you see the route in
Patrik Olsson wrote:
Silly me. Yes of course, if the route points to fxp0, the route will be
used for forwarding, but no packets can transit from the PFE there. But
Felix, are you sure packets go in the other direction?
yup. We've had it happen.
How does the connection between PFE and RE
Incidentally, I highly recommend placing a spoof-protect filter on your
fxp0 interface (something like: from source-address fxp0-network;
dest-addr fxp0-network; then accept; rest then reject), because all
packets entering fxp0 (e.g., broadcasts) with a non-fxp0-network
destination will be
Hi,
I want to know what is the default traffic update counter time which
updates ifOutOctets in ifTable for an interface, and how we can change it in
JUNOS8.1.
from our experience, JUNOS will update the snmp counter values every 5
seconds (if the system load is higher, it will take
Cord MacLeod wrote:
Below is a configuration for my ex4200. I'm trying to make it the
gateway for all my traffic temporarily. All devices connected to it are
running ospf, 2 m7s and another ex4200. All devices connected can hit
all other devices on loopback and /30 addresses. However, only
10.0.0.60/30*[OSPF/10] 07:09:00, metric 2
to 10.0.0.50 via ge-1/3/0.0
224.0.0.5/32 *[OSPF/10] 1d 23:32:36, metric 1
MultiRecv
On Jan 20, 2009, at 12:28 AM, Felix Schueren wrote:
Cord MacLeod wrote:
Below is a configuration for my
which version of JUNOS-EX did you use? We saw similiar effects with 9.1
and (to a lesser extent) with 9.2. Debugged for a good while, everything
100% correct, reboot, everything worked. We later learned that restart
ethernet-switching would probably have been sufficient.
There are quite a few
so far, multiple full tables, l2vpn, l3vpn,
flowSpec/flowFilter - everything we do on the M-series works on the MX
as well so far.
Kind regards,
Felix
--
Felix Schueren, Head of NOC
Host Europe GmbH - http://www.hosteurope.de
Welserstrasse 14 - D-51149 Koeln - Germany
Telefon: (0800) 4 67 83
to memory problems - the default m10i
came with 256MB RAM, and the box was very sluggish due to swapping,
which also logged entries like RPD_SCHEDULER_SLIP etc.
-felix
--
Felix Schueren, Head of NOC
Host Europe GmbH - http://www.hosteurope.de
Welserstraße 14 - D-51149 Köln - Germany
Telefon
51 matches
Mail list logo
