Hi Loernzo

No bug filed because I wasn't sure it was actually a bug. I figured it could 
just as easily have been something caused by other configuration changes I had 
made over time. To be sure it was a bug I'd need to first set up a box with 
factory settings and replicate the fault, and after the time I put in today 
getting my setup working, that further step fell into the 'too hard' bin.

I shouldn't be so lazy ;)

In answer to your second question; no and so hence the latter.

Pete
        

> On 27/04/2016, at 2:18 pm, Lorenzo Colitti <[email protected]> wrote:
> 
> Pete,
> 
> Did you file a bug on openwrt? My openwrt device here announces its ULA 
> address, but I think it does reply on that address.
> 
> Were you able to configure the ISP DNS servers using luci, or did you 
> hardcode them in the configuration file?
> 
> Cheers,
> Lorenzo
> 
> On Wed, Apr 27, 2016 at 11:11 AM, Pete Mundy <[email protected] 
> <mailto:[email protected]>> wrote:
> Dear ipv6-ops list members,
> 
> I just wanted to send a big props to you guys for the robust discussion on 
> the Marshmallow problem that Eric reported wrt slow web-page loading when v6 
> is enabled. It found myself bursting into laugher reading Tim's replies 
> (specifically the one including the phrase 'jammin on the jim jam frippen on 
> the krotz') and decided right there and then that bugger it, I need to get to 
> the bottom of the same problem that was happening here.
> 
> I had last week committed the cardinal sin of disabling v6 on my CPE (yes 
> barely a router, Jeroen) to get around exactly the same symptoms on my 
> Masrhmallow device. So today I spent no less than six (6!) hours digging down 
> to the bottom of it all using the combined knowledge I had gained, tcpdump, a 
> whole bunch of RTFM and many many binary tests.
> 
> To cut a long story short, my OpenWRT CPE was assigning it's LAN ULA v6 IP 
> via RDNSS, but it wasn't responding to any DNS queries sent to that address. 
> So DNS was borked for v6 queriers and the Marshmallow device wasn't elegantly 
> routing around the damage like all the other hosts on my (W)LAN were. A 
> similar run of events to those experienced by the ISP customers Eric reported.
> 
> The eventual work-around was to manually configure my ISP's (working :) v6 
> DNS resolver IPs into the odhcpd configuration.
> 
> So thanks everyone for your entertaining and informative discussions. Those, 
> coupled with a 1/2 day of my time have provided the pleasant outcome of 
> having v6 enabled again and all systems working :)
> 
> Pete Mundy
> Nelson, New Zealand

Attachment: smime.p7s
Description: S/MIME cryptographic signature

Reply via email to