> On Sep 24, 2016, at 7:47 AM, John Levine wrote:
>
>>> Well...by anycast, I meant BGP anycast, spreading the "target"
>>> geographically to a dozen or more well connected/peered origins. At that
>>> point, your ~600G DDoS might only be around
>>
>> anycast and tcp? the heck you say! :)
>
> Pe
As a Twitter network engineer (and the guy Patrick let camp out in your hotel
room all day) - thank you for this. Whoever was behind this just poked a
hornet’s nest.
“Govern yourselves accordingly”.
-C
(Obviously speaking for myself, not my employer…)
> On Oct 21, 2016, at 10:48 AM, Patrick
I have some experience with this; a few things off the top of my head:
- It’s usually best to leverage some sort of “smart” DNS to handle CNAME
distribution, giving you the ability to weight your CNAME distribution vs. only
using one CDN all the time, or prefer different CDNs in various global
And I’d *love* to hear the story they come up with when you ask why they only
want to rent space vs buy it…
-C
> On Jul 27, 2017, at 9:22 PM, Randy Bush wrote:
>
>> We were contacted by Admiral Hosting in London to rent some our
>> unused IP space.
>
> anyone wanting to rent/lease space is 99
Or..."Go ahead and keep buying 6509 chassis, the 7600 brand is just a marketing
thing"
-C
On Sep 14, 2011, at 7:41 AM, Leigh Porter wrote:
>
>
>> -Original Message-
>> From: Always Learning [mailto:na...@u61.u22.net]
>> Sent: 14 September 2011 14:39
>> To: N. Max Pierson
>> Cc: nanog@
Hearing rumblings of a major AT&T Wireless outage in southern California.
Anyone have more detail? Limited to cell towers or are transit circuits
affected?
-Chris
What Patrick said. For large sites that offer services in multiple data centers
on multiple IPs that can individually fail at any time, 300 seconds is actually
a bit on the long end.
-C
On Aug 18, 2012, at 3:43 PM, Patrick W. Gilmore wrote:
> On Aug 18, 2012, at 8:44, Jimmy Hess wrote:
>
>>
Dynect has a RESTful API as well. They even host a number of sample scripts at
GitHub:
http://dyn.com/managed-dns-dynect-5-api-access-load-balancing-geo-traffic-management/
https://github.com/dyninc
-C
On Sep 12, 2012, at 5:18 PM, Miles Fidelman wrote:
> Hi Folks,
>
> I expect folks on NANO
True, but it could be used as an alternative PMTUD algorithm - raise the
segment size and wait for the "I got this as fragments" option to show up...
Of course, this only works for IPv4. IPv6 users are SOL if something in the
middle is dropping ICMPv6.
-C
On Oct 29, 2012, at 4:02 PM, Templin,
Same reason no vendor has bothered to prune redundant RIB entries (i.e.
more-specific pointing to the same NH as a covering route) when programming the
TCAM...
-C
On Aug 13, 2014, at 1:42 PM, Randy Bush wrote:
> half the routing table is deagg crap. filter it.
>
> you mean your vendor won
>
> Pruning FIB entries, on the other hand, can be done quite safely as
> long as you're willing to accept the conversion of "null route" to
> "don't care." Some experiments were done on this in the IETF a couple
> years back. Draft-zhang-fibaggregation maybe? Savings of 30% in
> typical backbone
I've seen colos sell half-racks where both the top and bottoms of the racks
have their own cabinet doors. It's not a common thing though.
-C
> On Feb 12, 2016, at 18:58, Mike Hammett wrote:
>
> There are more options when you're not just using someone else's datacenter.
>
>
>
>
> -
>
I think that’s the problem in a nutshell…until every vendor agrees on the size
of a “jumbo” packet/frame (and as such, allows that size to be set with a
non-numerical configuration flag). As is, every vendor has a default that
results in 1500-byte IP MTU, but changing that requires entering a va
OK, this sounds Really Wacky (or, Really Hacky if you're into puns) but there's
a reason for it, I swear...
Will typical OSS UNIX kernels (Linux, BSD, MacOS X, etc) reply to a crafted ARP
request that, instead of having FF:FF:FF:FF:FF:FF as its destination MAC
address, is instead sent to the al
viously can't read RFCs, the answer
is "yes". :)
-C
On Jun 16, 2010, at 3:57 51PM, Chris Woodfield wrote:
> OK, this sounds Really Wacky (or, Really Hacky if you're into puns) but
> there's a reason for it, I swear...
>
> Will typical OSS UNIX kernels (Linu
So let us commence the shipping of stupidly overpriced silicon...802.3ba is an
official IEEE standard.
http://www.businesswire.com/portal/site/home/permalink/?ndmViewId=news_view&newsId=20100621006382&newsLang=en
-C
The unconfirmed chatter I'm hearing is that they were leaking peering routes to
other peers. Can anyone check and confirm this? Renesys?
-C
On Sep 16, 2010, at 9:09 12AM, William Byrd wrote:
> XO Engineers are telling us that they are aware of packet loss across their
> network and are looking
On Sep 16, 2010, at 10:57 07AM, George Bonser wrote:
>>
>
> Hi Chris,
>
> Since prioritization would work ONLY when the link us saturated
> (congested), without it, nothing is going to work well, not your
> torrents, not your email, not your browsing. By prioritizing the
> traffic, the torrents
On Sep 17, 2010, at 6:48 02AM, Jack Bates wrote:
> On 9/17/2010 4:52 AM, Nathan Eisenberg wrote:
>>> True net-neutrality means no provider can have a better service than
>>> another.
>>
>> This statement is not true - or at least, I am not convinced of its truth.
>> True net neutrality means
On Sep 17, 2010, at 9:23 09AM, Jack Bates wrote:
>
> Is it unfair that I pay streaming sites to get more/earlier video feeds over
> the free users? I still have to deal with advertisements in some cases, which
> generates the primary revenue for the streaming site. Why shouldn't a content
> pr
Agreed; my reading of this suggests database caching issues (i.e. all the
frontend/middleware clients hitting the main sql cluster at once instead of the
memcached farm they normally use), not HTTP/CDN caching issues.
-C
On Sep 23, 2010, at 7:17 12PM, Jay R. Ashworth wrote:
> http://www.facebo
Historically, you would find that routers designed for long-haul transport
(Cisco GSR/CRS, Juniper M-series, etc) generally had deeper buffers per-port
and more robust QoS capabilities than datacenter routers that were effectively
switches with Layer 3 logic bolted on (*coughMSFCcough*). That li
I know of one large-ish provider that does it exactly like that - RIPv2 between
POP edge routers and provider-managed CPE. In addition to the simplicity, it
lets them filter routes at redistribution without having to fiddle with
inter-area OSPF (or, ghod forbid, multiple OSPF processes redistrib
On Sep 29, 2010, at 6:14 PM, Scott Morris wrote:
> But anything, ask why you are using it. To exchange routes, yes... but
> how many. Is sending those every 30 seconds good? Sure, tweak it. But
> are you gaining anything over static routes?
For simple networks, RIP(v2, mind you) works fine.
Ability to route IPv6 != ability to route IPv6 as well as IPv4. Depending on
the hardware, there will always be unavoidable tradeoffs, which tend to be
either in reduced throughput capacity, typically noticed on particularly on
software-switching platforms, or the number of routes/ACLs/etc you c
What's interesting is that the !NANOG part of the universe presumes
the maintenance was to be performed by Twitter, not by their carrier
(i.e. server, not network, upgrades). Given the fact that the
WhaleFail has become a commonly-recognizable sight, I can see this
make people a bit, um, ne
To pile on in the spirit of "if people don't complain, nothing will
change" - is VZB still insisting on filtering >/32 at their peers?
While ARIN is allocating /40s and /48s directly?
-C
On Mar 10, 2010, at 2:18 PM, Seth Mattinen wrote:
On 3/10/10 11:00 AM, Charles Mills wrote:
Does anyon
Hi,
Was wondering where one in the SF Bay area might be able to borrow (or
otherwise procure at a reasonable cost) a short - less than 1 meter - section
of undersea fiber cable for a presentation I'll be giving in a few weeks. Feel
free to unicast your reply if you are in a position to assist.
(Yeah, high reply latency...)
Is Carrier V still filtering at sub-/32 on their IPv6 peerings? Last I was in a
position to check, not even Apple's /45 was visible from inside AS701.
-C
On Feb 10, 2011, at 12:25 PM, Eric Clark wrote:
> Don't remember about the v4 part, but 3 years ago they issue
I think this is the point where I get a shovel, a bullwhip and head over to the
horse graveyard that is CAM optimization...
-C
On Mar 8, 2011, at 5:18 20PM, Chris Enger wrote:
> Our Brocade reps pointed us to the CER 2000 series, and they can do up to
> 512k v4 or up to 128k v6. With other Br
30 matches
Mail list logo
