neighbor a.b.c.d allowas-in route-map SAFETY
Wow this would be so cool, I'll definitely mention this to our SE.
I was wondering if the internet service is realized as MPLS VRF than the ISP
could do as-override which is pretty standard for VPN services.
What I'm curious about is the percentage
A BGP speaker will not accept routes with its own AS in the path, nor should
it.
Whilst a number of people have suggested using allowas-in, I'd personally not
recommend it as loop prevention is generally a good thing - if you ever added
a BGP speaker to one network that also had internet
Date: Mon, 17 Jun 2013 22:04:52 -0600
From: Phil Fagan philfa...@gmail.com
... you could always
thread the crap out of whatever it is your transactioning across the link
to make up for TCP's jackknifes...
What is a TCP jackknife?
Cheers.
Jakob.
It is also called a sawtooth or similar terms. Just google tcp
sawtooth and you will see many references, and images that depict the
traffic pattern.
HTH,
Fred Reimer | Secure Network Solutions Architect
Presidio | www.presidio.com http://www.presidio.com/
3250 W. Commercial Blvd Suite 360,
Thanks Fred. Sawtooth is more familiar.
How much of that do you actually see in practice?
Cheers,
Jakob.
On Jun 18, 2013, at 6:27 AM, Fred Reimer frei...@freimer.org wrote:
It is also called a sawtooth or similar terms. Just google tcp
sawtooth and you will see many references, and images
Sorry; yes Sawtooth is the more accurate term. I see this on a daily
occurance with large data-set transfers; generally if the data-set is large
multiples of the initial window. I've never tested medium latency(
100ms) with small enough payloads where it may pay-off threading out many
thousands of
Dear All
We Deal with TCP window size all day every day across the southern cross from
LA to Australia which adds around 160ms... I've given up looking for a
solution to get around physical physics of sending TCP traffic a long distance
at a high speed
UDP traffic however comes in very
The Skid-mounted C3Spear in a NEMA4 version would work perfectly for this
application.
http://www.ellipticalmobilesolutions.com/c3spear.html
So I'm clear, its not just a low bit rate argument. Its a low bit rate,
combined with little spare horsepower (CPU RAM), little non-volatile
storage, and a deluge of information to sort through in order to find
something useful. If core routers weren't in the core, where they have
access to
On Tue, 18 Jun 2013 15:53:48 -, James Braunegg said:
We Deal with TCP window size all day every day across the southern cross from
LA to Australia which adds around 160ms... I've given up looking for a
solution to get around physical physics of sending TCP traffic a long distance
at a
On Mon, Jun 17, 2013 at 7:38 AM, Tony Patti t...@swalter.com wrote:
Thanks, I liked your pointer to the SDR.
But can I ask you for a bit more info about your statement
where oscilloscopes and frequency analysis is available to anyone with some
Google-fu
We don't need as much test equipment
now THAT would be a cool project!
On Tue, Jun 18, 2013 at 2:25 PM, Jazz Kenny trapperjohn...@gmail.comwrote:
On Mon, Jun 17, 2013 at 7:38 AM, Tony Patti t...@swalter.com wrote:
Thanks, I liked your pointer to the SDR.
But can I ask you for a bit more info about your statement
where
There's already code out there for the GNURadio project's software defined
radio infrastructure that supports some very basic LTE analysis using a
$20 or less USB DTV tuner stick!! Only a matter of time before some radio
devices with a lot more bandwidth become affordable and easily
On Tue, Jun 18, 2013 at 02:31:37PM -0600, Phil Fagan wrote:
now THAT would be a cool project!
(I missed the beginnig of this thread; sorry if this is a repeat.)
There was the fellow demonstrating a spoofed 2G GSM tower at DefCon
recently:
Howdy,
I have been working on a proposal for the organization I work
for to move into the 10gbit datacenter. We have a small datacenter currently
of about 1000 ports of 1gbit. We have traditionally been a full Cisco shop,
however I was asked to do a price comparison as well as
I love JUNOS, don't really care for IOS. I really trust Cisco and Juniper's
hardware, with that being said Arista is your best bet for cheapest port.
I've only seen Arista in lab, not in the wild yet so I can't speak for how
I would trust them. You mention getting bit by single sups, I believe as
Dear Valdis
Thanks for your comments, whilst I know you can optimize servers for TCP
windowing I was more talking about network backhaul where you don't have
control over the server sending the traffic.
ie backhauling IP transit over the southern cross cable system
Kindest Regards
James
Let me also clarify, Price per port is not the final deciding factor. We are
looking much more at a combination of daily operational sanity, troubleshooting
features, operational feature set, vendor support quality and price.
Support is absolute key. When we need help, we need help quickly
DELL Force10 switches (not DELL Power Connect) run so far so good in
our environment. the combination of S4810 and Z9000 make good sense on
both operation and capex point of view.
There were three headaches for us in the beginning of adaption.
Force10 calculates frame size with CRC32, say if your
On Wed, 19 Jun 2013 00:24:15 -, James Braunegg said:
Thanks for your comments, whilst I know you can optimize servers for TCP
windowing I was more talking about network backhaul where you don't have
control over the server sending the traffic.
If you don't have control over the server,
On Tue, Jun 18, 2013 at 5:31 PM, Blake Pfankuch - Mailing List
blake.mailingl...@pfankuch.me wrote:
Let me also clarify, Price per port is not the final deciding factor. We are
looking much more at a combination of daily operational sanity,
troubleshooting features, operational feature set,
On Tue, Jun 18, 2013 at 08:47:41PM -0400, valdis.kletni...@vt.edu wrote:
On Wed, 19 Jun 2013 00:24:15 -, James Braunegg said:
Thanks for your comments, whilst I know you can optimize servers for TCP
windowing I was more talking about network backhaul where you don't have
control over
I've had nothing but good luck with Juniper support and well with Cisco you
pay for support too. I will say Arista support was great, however, I'm
still hesitant to put them in full production; but I think that is lack of
experience with them speaking.
Do the bake off in your lab and let'm run!
I'm exact opposite of Phil. I love IOS and hate JunOSfor that
single reason, I'm really against buying Juniper in our shop for
pretty much anything. :)
Still, to be fair, the hardware seems to be really, really stable and
well built. I don't think we've had a failure across our Junipers in
Mike brings up a good point though; the effort, cost, and risk of
introducing a new CLI to an environment sometimes is masked until you
really need to dig in and work through outages. Familiarity with a codebase
or at least with how the code thinks should go a long way when deciding
what to put in
On Tue, Jun 18, 2013 at 4:53 PM, Blake Pfankuch - Mailing List
blake.mailingl...@pfankuch.me wrote:
Howdy,
I have been working on a proposal for the organization I
work for to move into the 10gbit datacenter. We have a small datacenter
currently of about 1000 ports of 1gbit.
Go juniper!!!
Full junos equipment on the network means same OS for switches, routers,
and firewalls.
You have high end equipment to support a core tier1 backbone, and also a
simpliest 24 port sw soho range. All with the same config languaje.
You can use the management software called junos space
27 matches
Mail list logo
