netflix proxy/unblocker false detection

Owen DeLong owen at delong.com
Fri Jun 26 13:35:58 UTC 2020


I take his statement more as:

	“If Netflix wasn’t doing IPv6, they’d be in more of a corner
	to resolve CGNAT issues. Since they support IPv6, likely their
	response to CGNAT issues is ``Press your provider to do IPv6,
	it’s better.’’”

Likely, that is true. Support for IPv6 isn’t at fault here. Rather, the
reality that IPv6 is a relatively easy way to offer a much better user
experience than CGNAT is in play here.

Owen


> On Jun 25, 2020, at 7:45 AM, Christian <cdel at firsthand.net> wrote:
> 
> wow. blaming support for IPv6 rather than using cgnat is a huge stretch of credibility
> 
> On 25/06/2020 10:20, Mark Tinka wrote:
>> 
>> On 25/Jun/20 11:08, Denys Fedoryshchenko wrote:
>> 
>>> Did anybody noticed that Netflix just became useless due to tons of
>>> proxy/unblocker false detection on CGNAT ranges?
>>> Even my home network is dual stack, i am absolutely sure there is no
>>> proxy/vpn/whatsoever (but ipv4 part is over CGNAT) - and i got
>>> "proxy/unblocker" message on my personal TV.
>>> And many other ISP sysadmins told me that recently this is a massive
>>> problem, and netflix support is frankly inadequate and does not want
>>> to solve the problem.
>>> I will not be surprised that they will begin to actively lose users
>>> due to such a shameful silly screwed up algorithm.
>>> Who in sober mind blocks all legit users due probably one or two
>>> suspicious users behind same IP range?
>> This isn't a new problem - for years, services that track what a single
>> IP address does can deny access if something looks amiss.
>> 
>> Of course, CG-NAT is a reality, but perhaps Netflix find it will be
>> easier to lose some customers than building infrastructure and support
>> to work out what is valid CG-NAT vs. mischief.
>> 
>> Probably would have been an easier case if Netflix didn't support IPv6,
>> but alas...
>> 
>> Mark.
> 
> -- 
> Christian de Larrinaga
> ----------------------



More information about the NANOG mailing list