Looks like Be and O2 are having more routing problems – this time affecting traffic to OpenDNS servers and some others. Things have stopped working several times on both Be and O2 connections I use in London over the last 24 hours, sometimes for a few seconds, sometimes for a few hours. Advice from both providers is to set router DNS to the original factory settings if you’ve changed them. Thing is, they’re offering this advice in their online forums, and uh… well without working DNS servers to look at, affected users won’t be able to get to the forums in the first place.
So on getting up this morning I found that I had no Internet (that is, web, mail or other IP) connectivity to any site outside of the UK, and even then the UK-based sites were patchy at best. I thought at first that someone’s nearby WiFi box had switched channels again, but even over a cable things were still just as bad. I kept an eye on the Be forums and discovered that this was a nationwide problem, with LINKS being blamed again by many. As yet there’s no online confirmation of anything being fixed, but the connection reliability has been improving throughout the morning.
This the third such major outage I’ve experienced with Be/O2 over the last year, and it’s getting quite tiring. Makes me wonder whether it’s time to change ISP again?
So it’s been a day or two now since Be did some works on their infrastructure and changed the status of their “Pages loading half way” forum thread to “fixed”.
Well, it doesn’t seem to be from where I’m sitting. I’m still seeing anything between 4 and 10% packet loss to any web server, and pages requiring reloads more often than usual. Others are noticing the same thing, and some on their user forums have even accused Be of simply brushing the issue under the carpet.
It’s clear there have been some significant peering issues affecting Be (and other ISP’s in the UK) over the last few months, and there seems to be a rather large group of customers who feel they’re not getting the service they’re paying for, because they’re not seeing the improvement they’re asking for – and to rub salt into their wounds they’re not seeing the ISP’s raise the issue successfully outside of their own organisations.
Whatever others feel, for me it would be nice to have some acknowledgement other than “we’re looking at this, please send us any info you can”. It might just be a longer-term passing glitch. Here’s hoping.
For the last couple of days I’ve been noticing that web pages on mine or my wife’s Macs sometimes do not load completely. I’m using Chrome and my wife uses Safari, which I think rules any browser-related issue. I’ve noticed the same issue at work, so I ran some Ping tests and I’m finding something like 2-4% packet loss between our London office location and the BBC News website. Usually I’d expect this to be 0%. No wonder some things aren’t loading properly.
The issue occurs regardless of whether we’re using WiFi or cable to connect to the router. Given that O2 and Be share some backhaul for ADSL2 services I’m not surprised that I’m seeing the same issue at home (O2 24Mb) and at work (Be “Pro” 24Mb). What did surprise me was trying our BT Business Broadband 8Mb ADSL connection and finding the same issue, with similar frequency.
A thread appeared on the Be customer forums yesterday, but I cannot link to it here as it needs login credentials. What I can link to is a related post on the Be unofficial usergroup, explaining that “Strange Things” are ongoing – affecting traffic to random websites at random intervals.
I’m wondering if this is a common experience, and whether anyone out there knows what the problem is?
Anyone know why WiFi on these boxes stops transmitting Bonjour data, thus breaking Time Machine, Remote Desktop Sharing, AFP file sharing among others? It seems to correlate with signal quality and/or interference from neighbouring networks. It seems that once every week or so the quality goes down enough that network traffic grinds to a halt over WiFi. Changing the WiFi channel frequency to one that is less congested seems to help for another week.
The issue affects our G4 iMac (OS X 10.4), 2007 MacBook Black (OS X 10.6) and 2009 MacBook Pro 13″ (OS X 10.6) uniformly. Don’t have any Windows or Linux boxen to compare service. Our iPhones don’t use Bonjour for anything I can think of, so I don’t know whether the issue would affect them too.
Connecting to the same router/network using a cable seems to always work, so it is something specific to the WiFi section of the modem.
I’m confused and would like the nonsense to stop… Can anyone help?
Ugh. Just been struggling all afternoon with a printing issue while working from home. All I wanted to do was to print one test chart – then lost an hour or more trying to fix it, until I happened to find a fix online – thank you Uncle Google.
The problem is this – my macs use Bonjour (a Zeroconf service) to communicate what they are able to share with each other on my network. This works well on cabled links, but for some reason fell apart whenever I use WiFi – which frankly is all the time as I can’t be bothered with having a cable trailing across the room whenever I want to surf.
REASON AND WORKAROUND
For reasons unknown, It appears the O2 Wireless Box III does not correctly/fully bridge all protocols between WiFi and Ethernet-based connections, Bonjour included – but only when the WiFi is configured to use N-mode. Configure the box to use either of the the slower B or G modes and my world is a happy place.
So yet again I’m having to sacrifice speed for reliability. Does nobody test these things? One would think my usage scenario is not uncommon, and it turns out that users across the UK are stuck with this issue, both with the II and III versions of the box.
Still, the box was free, so I guess I should count my blessings…