[kwlug-disc] [kwlug-announce] Meeting Monday: OpenWRT

Paul Gallaway paul at gallaway.ca
Thu Aug 14 14:19:23 EDT 2014


On Thu, Aug 14, 2014 at 11:26 AM, Khalid Baheyeldin <kb at 2bits.com> wrote:
> That one is very similar to my DLink DIR-835, but with 8MB flash instead of
> 16MB, and making up for it by having 2 USB ports not just one like mine.
Looking at the presentation lafter getting the router home I noticed
you had the DLink which had MORE flash than the WDR4300 I had a doh!
moment until I found out it was not in stock at Canada Computers
(although also on-sale this week). I'm happy to have something that
can work now and hopefully it's not obsoleted too quickly. 16MB in the
DLink should be more future proof for now, but with the high end
routers coming with 128MB of flash now, that might not even be enough.

> Saw the APC BackUPS on Dell's site yesterday for $45 after a $10 off.
That's the one.

> Just curious how much was the gigabit hub?

$16 unmanged 8-ports. Description says it's the metal version but the
photo and model number are the plastic one. I'm expecting it to be the
plastic one. I was trying to not spam the list with 'sales'
information but since you asked!:
https://accessories.dell.com/sna/products/Networking/productdetail.aspx?c=ca&l=en&s=dhs&cs=cadhs1&sku=A7312374

> If you ever do sysupgrade, please post a followup on the steps you did
> (sysupgrade of a xxx-sysupgrade.bin file, installing packages via opkg
> again).

Will do. I'm not there yet :)

> I am deferring doing the upgrade until a stable release comes out .

Everything I read indicated that there were some issues installing
A.A. on the Version 1.7 hardware that I have but a work around had
been reported to work for B.B. with the possibility of it working on
A.A. So I took the path that reported it definitely worked by
installing the most recent release candidate for B.B. (RC3 at time of
writing). I was pleased to find that everything just worked and the
graphical interface came right up and the work around was unnecessary.
I'll need to do some testing before I deploy.

> My DLink DIR-835 is in the basement (very close to the roof, as high as it
> can get), and the first floor above it gets good coverage. The coverage in
> the 2nd floor is not as good but still acceptable. This is on Wireless N
> 2.4GHz.

Same here. Currently in the ceiling of my basement. I started thinking
about better places to put it and I think I could do some wire drops
and put the router in a better spot. Any downside to moving the router
15 ft from the modem (different floors) or should I move the modem
too? I'm planning to run it alongside the current router for a week or
so (hopefully starting this weekend) before replacing the old one.

On Thu, Aug 14, 2014 at 1:38 PM, CrankyOldBugger
<crankyoldbugger at gmail.com> wrote:
> Not to rain on the parade here... but I noticed a number of people
> recommending TP-Link.  I seem to recall several months ago (maybe over a
> year ago) a big stink about TP-Link routers having backdoors.  Here's one
> such example (looking for more):
> http://malwarelist.net/2013/03/15/backdoor-in-tp-link-devices/

I am aware of this (or maybe an issue like this) and I would argue
that this /could/ apply to any OEM firmware that comes in binary form
and also why OpenWRT went from something I thought would be nice to
have, to something I felt was a necessary feature. I think this type
of vulnerability could exist for any hardware that uses a binary blob
for its driver. I remember reading something on Ars:
http://arstechnica.com/security/2013/01/secret-backdoors-found-in-firewall-vpn-gear-from-barracuda-networks/

How many other manufactures have a similar backdoor in their hardware?
How could we ever know about them all? I seem to recall a follow-up
article stating that the backdoor had been fixed according to the
vendor but hadn't been when tested.

> My question is.. would openWRT negate the backdoor, or is the problem a bit
> deeper in the router?

I think it would negate the issue, and seeing as the router I just
bought is on that list I have a vested interested in confirming it is
not vulnerable. My thought is that since OpenWRT is compiled from
sources using modules supplied by the linux kernel, it's about as safe
as it can ever be made. That's why I choose OpenWRT over DD-WRT,
Tomato, and all the other WRT variants. My understanding of the
inner-workings of the hardware and how the kernel interacts with all
the bits of silicon in the device is limited, so if I'm flat out
wrong, someone please tell me!

On Thu, Aug 14, 2014 at 2:05 PM, Khalid Baheyeldin <kb at 2bits.com> wrote:
> Those who have that router can test using the proof of concept that is
> detailed here
>
> http://sekurak.pl/tp-link-httptftp-backdoor/

Thanks Khalid. Haven't checked out the link yet but I'll see if I have
the capabilities to test and report back the results. If I can't test
it I will report back looking for help ;).





More information about the kwlug-disc mailing list