What else could it be? - Page 5

Welshchris · 27-10-2009, 17:29

Local cabling, there r 2 others affected that are atleast 5 miles away from i am i wouldnt call that local cabling!

---------- Post added at 18:29 ---------- Previous post was at 18:25 ----------

Quote:

Originally Posted by Broadbandings

That'll point to a fault with your local cabling in that case, assuming that everyone is dropping out simultaneously with you. Certainly doesn't indicate DHCP though, that'd kill everyone in the entire area off as their leases expired.

Given that you don't believe that DHCP is ok why are you so sure that the tech guys are correct saying that the cabling is fine, especially in light of how you originally reported this, as spiking upstream transmit power?

You do know that the local network engineers have absolutely no way of checking the 'spanning tree' on the DHCP servers, right?

Did you ever capture a report of how long one of the leases that the DHCP is giving you is?

I am only going with an email that i was sent by a second line support agent and a network engineer in swansea.

I have captured the leases and it tends to die around the time the leases are up, im never given the same lease time, its either 24hrs, 48hrs, 6 or 7days.

Ignitionnet · 27-10-2009, 17:35

Quote:

Originally Posted by Welshchris

Local cabling, there r 2 others affected that are atleast 5 miles away from i am i wouldnt call that local cabling![COLOR="Silver"]

People 5 miles away are on the same BSR card as you?

I think you have to be mistaken here, they are unlikely to cover that wide an area, an entire legacy uBR, all 4 cards, covers an area similar to or wider than a BSR / 10k card.

Again if this were affecting everyone in Swansea due to being a DHCP server tech support agents would be jumping up and down to get it fixed and it would be a P1 or MSO. It would be very obvious that this is an issue as fault levels in Swansea would go through the roof.

Have second line told you, at any time, that it's the DHCP server at fault? Local network guys have no access to either that DHCP server or the switch it is connected to and have no way to conclude that it is what is at fault.

Regarding the repull - if you are so certain that it is not what is at fault it may be wasteful but it's what happens when you sent flurries of emails and letters of complaint to every man and their dog, a number of different people going in different directions to try and deal with the same issue. Too many cooks, broth, etc.

If you are sure the repull isn't necessary you can refuse it. Sounds like VM are as confused as you are on this issue.

Welshchris · 27-10-2009, 17:48

check a map Im in high street in swansea, one of the others here are in Fforestfach in swansea, other is in Cwmburla where the UBR/BSRs are and they have sent me messages with what cards they r on and they r on identical to me, the only thing that differs is customer number....

Local network team have been in touch not only with second line support but also a team above them which has told them its a possible spanning tree error on the BSR as its been witnessed with a few others.

Here are all the times of drop outs sine October 14th....
Bare in mind each time it drops the DCHP gives me a new IP lease also and in the event viewer it tells me that it lost its lease from Modems IP address...

14th Oct..
16:21 + 17:03

15th Oct..
DATE GIVEN 50mb...20:51 + 21:56

16th Oct..
16:11

17th Oct..
00:48

18th Oct..
4:03am
4:24am
Switched back to 20mb as i was told to use for 3 days only...

19th Oct..
4:09am
16:00
19:24
Told by engineer who found someone with identical issues to me to switch back to 50mb so diagnostics could be run.

20th Oct..
00:15
00:20
19:26
19:28

21st Oct..
20:21
20:30

22nd Oct
4:49am
18:06
18:10

23rd Oct..
13:05 and back to 20mb as told...

24th Oct..
1am and couldnt get an IP address until 5:44am

13:10
15:32
16:10
16:29 and switched back to 50mb cos i lost temper to see if it would make any difference...

19:45

25th Oct..
4am
4:07am
16:26
16:30
21:14
22:34

26th Oct..
11:24am
14:18 and back to 20mb
16:30

27th Oct
12:05
12:08
12:37
12:49
12:55

and thats so far today.

Ignitionnet · 27-10-2009, 18:15

I have my doubts that Virgin are giving leases lasting 3 minutes or 12 minutes. The shortest possible lease on the network which was as far as I know for pre-registration modems only, is 24 hours. You are losing sync due to RF issues.

Your PCs losing their IP addresses will not cause the modem to go offline. The modem is not the device handing out IP addresses and if the PCs are claiming to have lost their lease from the modem they are wrong, the modem is not visible and the IP lease should be sourced from the DHCP server not the modem. It is impossible for the modem to give a public IP address lease.

Of course, if your devices think they are losing lease from the modem, where do you think they are going when they try to renew it. No wonder it's failing if they are harassing the modem to renew a lease it never supplied in the first place, they should be request the renew from VM's DHCP servers.

If you have a PC directly connected to the modem and you do an ipconfig /all it should show a VM DHCP server, if it's showing 192.168.100.1 this is the modem and is wrong. The modem's IP address is for management or when it loses sync in which case it may give an IP address to devices behind it which is probably what happens from time to time.

The above scenario would also explain why a modem reboot is required. Until ethernet goes down for a while your kit keeps harassing whatever it can find for an IP address.

EDIT: You've been in IT for 12 years apparently, if you haven't just spent a decade plugging DIMMs and cards in you should be good to get Wireshark running on a PC connected directly to a modem and capture some traces of the DHCP transactions.

Welshchris · 27-10-2009, 18:20

So is there anything that u can suggest i can tell them?

Oh and btw i know ur goin to tell me this isnt possible but this was witnessed by a Virgin tech here also that couldnt believe it also....

i was issued a 169ip address and was still able to go online for over 4mins.

Ignitionnet · 27-10-2009, 18:22

Quote:

Originally Posted by Welshchris

So is there anything that u can suggest i can tell them?

Once your modem stops falling offline I suspect your IP address issues will also go away.

If you can get Wireshark running it will supply some more information on the DHCP traffic. I think what's happening is that your kit is trying to renew DHCP lease using the modem which doesn't work too hotly. I'm not sure why or how it gets the first initial lease but suspect it's trying to renew from the modem.

You can check this by noting the output on ipconfig /all on a PC connected directly to a modem.

And no I'm not going to disbelieve the 169 thing at all, appearances can be deceptive and I've seen what looks at first glance like weird stuff but actually made sense in the end.

Welshchris · 27-10-2009, 18:29

i have it running and will keep it running until the next drop

Ignitionnet · 27-10-2009, 18:33

Enter this as a filter:

ip.addr == 192.168.100.1

Should show if it's sending anything untoward to the modem.

Go to a command prompt and type this:

ipconfig /all | find "DHCP Server"

See what output you get from that, copy/paste it here.

Welshchris · 27-10-2009, 19:11

i have no idea how to use this software what so ever.

---------- Post added at 20:11 ---------- Previous post was at 19:46 ----------

when applying the filter theres no traffic what so ever been shown

Welshchris · 29-10-2009, 13:09

Some shots from Wireshark on the problem!

justice · 30-10-2009, 18:09

Hi chris just want to say have had the exact same problem for the last 8-12 weeks easy i am also in SA5. modem reboots every now and again sometimes sevral times in a row. have tried ringing VM but just dont like to wait for hours to get through to be told the same thing over and over . also been with cable since the start many many years ago this is only the second problem ive had lol how ironic . funny tho the problem coincided with the replacement of a large cabinet on the corner of pentrigething and fforydd brain outside the council offices there after the old one was demolished by a car . have no idea if revelant but still thought it was stange. hope your problem sorted soon because have a funny feeling that mine will be at the same time.

Welshchris · 30-10-2009, 19:30

Justice there r atleast 5 of us now that im aware of in Swansea with itentical issues.

justice · 30-10-2009, 21:47

i don't doubt that m8 thought i was the only one must be a real big reason then. bet theres more that just brush it aside like me. Not Right tho bloody anoying sometimes

Ignitionnet · 30-10-2009, 22:33

Quote:

Originally Posted by justice

funny tho the problem coincided with the replacement of a large cabinet on the corner of pentrigething and fforydd brain outside the council offices there after the old one was demolished by a car.

Sounds a tad more likely than random ramblings regarding servers local networks staff cannot monitor that serve wide areas of tens of thousands of customers that appear largely unaffected from a data centre in Swansea.