[OT] Re: wireless routers
lesleyb at herlug.org.uk
Tue Nov 9 17:52:21 GMT 2010
On Tue, Nov 09, 2010 at 01:34:43PM +0000, Chris Jack wrote:
> On Mon, Nov 8, 2010 at 2:08 PM, Dirk Koopman <djk at tobit.co.uk> wrote:
> > My netgear WPN802 seems to have died. Any recommendations for a non-ADSL
> > wireless router that is reliable and might last a bit longer? Don't need
> > anything fancy, but I would like something that I don't have to reboot
> > regularly.
> I've always wondered a bit about this rebooting thing - whether the real problem is at my end or that of my service provider. I just have seen so many people buy new routers with the expectation they won't have to reboot - only to find themselves having to reboot the replacement.
> I also wonder if the service provider is doing something to intentionally cripple people who have been generating a lot of traffic for a period of time and whether rebooting is really just effectively resetting the counter at the service provider's end.
> I use Sky - and on a semi-regular basis it stops working Friday night or the weekend (regardless of what I reboot) - and then miraculously springs back into life on Monday morning. My take on this has been that Sky doesn't employ engineers to do weekend call outs to fix things that break their end.
> Not so long ago, my Sky box stopped working even mid-week - lights seemed to sort of come on but no connection and no LAN (which definitely wouldn't be Sky's end). Interestingly Sky's first suggestion was to send us a new power cable which I thought would have little chance of success (as lights were coming on on the router al beit not as enthusiastically as they had when it worked) - but surprisingly it fixed the problem.
I have attached a venerable ZyXel 660HW to a talktalk connection - and it sends me logs so I can see what happens.
Things seem to have improved ... ish. I was getting multiple drop outs per day but these events seem to be less frequent with an occasional burst to remind me what a pig this cheap connection can be.
The connection gets dropped - call terminated occurs in the logs . After a short delay an outgoing call is made a packet trigger either sent or received and a connection is logged.
The ppp:LCP connection process then starts and the call gets terminated again. Here's some logs I prepared earlier ;).
18|04/12/2010 00:49:17 | | | ppp:LCP Starting
19|04/12/2010 00:49:17 | | |CALL DETAIL RECORD board 0 line 0 channel 0, call 14, C02 OutCall Connected 512000
20|04/12/2010 00:49:17 | | |PACKET TRIGGER Packet Trigger: Protocol=1, Data=Packet Trigger: Protocol=1, Dat
21|04/12/2010 00:49:17 | | |CALL DETAIL RECORD board 0 line 0 channel 0, call 14, C01 Outgoing Call dev=5 ch=0
22|04/12/2010 00:49:17 | | |CALL DETAIL RECORD board 0 line 0 channel 0, call 13, C02 Call Terminated
23|04/12/2010 00:48:17 | | | ppp:LCP Starting
24|04/12/2010 00:48:17 | | |CALL DETAIL RECORD board 0 line 0 channel 0, call 13, C02 OutCall Connected 512000
25|04/12/2010 00:48:17 | | |PACKET TRIGGER Packet Trigger: Protocol=1, Data=Packet Trigger: Protocol=1, Dat
26|04/12/2010 00:48:17 | | |CALL DETAIL RECORD board 0 line 0 channel 0, call 13, C01 Outgoing Call dev=5 ch=0
27|04/12/2010 00:48:17 | | |CALL DETAIL RECORD board 0 line 0 channel 0, call 12, C02 Call Terminated
Eventually the thing progresses to the CHAP and IPCP stages but sometimes it gets stuck in a loop, connecting and disconnecting.
I try a soft reboot, a hard reboot and then an 'ard reboot with interim complete isolation of the router in both power and ADSL lines.
One of these has always cleared the fault.
Talk Talk would probably immediately blame my router if I complained and I would lose observation of the connection if I installed theirs.
You are probably right to suspect Sky doesn't repair stuff at their end over the weekend and I have no idea if you can soft reboot or are restricted to hard reboots on your box but Sky should be able to provide you a service over the weekend.
Curious about the power cable though. If it is a plain old power cable there isn't much a power cable can do to fail a piece of equipment besides die completely so I would doubt the power cable itself made the difference. Was it a different length or did it have post-transformer ferrite core screening?
More information about the london.pm