PDA

View Full Version : Odd ping times....


DarthMuppet
18-05-2010, 19:47
started getting problems last tuesday, 25% of pings going astray to bbc.co.uk, hp.com, google.com and primary DNS - always at the same time for all servers. Engineer cam on Thursday at my request, re-terminated everything, and i got him to do the 50MB upgrade whilst he was there (he had the kit and got approval after i coughed up). now seems to be faster obviously, 50MB easily achievable, bit the ping results are still no better. I work in IT for a large corporate, and 10% loss of *any* traffic, never mind ICMP would get me a major ass-whupping in a service review, but it seems to be the norm here.

At this point, the modem now directly connected to the PC. I have an MTU of 1300, not that it is relevant as all 4 machines show the same thing at the same time whether they are connected to the modem or via the router.

I've just come back from dinner to find the last 20 minutes have been entirely free of ping responses. I had to to release and renew my IP address to get reconnected.

So - is ~20% of ICMP loss normal, or is there something else amiss here? I've never really paid that much attention, as I very rarely have problems. I've spent the last 20 weeks basically working from home full time, and i've only had a problem since last week.

TIA
DM

Chrysalis
18-05-2010, 20:59
well the mtu for a start seems too low. 1500 should work fine.

if your net went dead for that 20 minutes, it may be your modem is having trouble staying connected, I suggest you post the modem stats and let someone like seph and ignition diagnose it.

Sephiroth
18-05-2010, 21:19
well the mtu for a start seems too low. 1500 should work fine.

if your net went dead for that 20 minutes, it may be your modem is having trouble staying connected, I suggest you post the modem stats and let someone like seph and ignition diagnose it.

In that case, please post all the downstream, upstream and event log information (masking out your CM-MAC address). Even better if you can do it directly conencted to the cable modem.

Please also do PATHPING www.bbc.co.uk and post the full results. Pathping takes quite a few minutes to run and looks as if it's hanging and it isn't.

As you prolly know, ICMP packets are discardable and the important thing to look for in the pathping is that there is no packet loss in transit to the destination node.

Cheers

sparksy
19-05-2010, 12:37
I am having similar problems - a few times per minute I am getting total packet loss to all destinations. I can see this when my keystrokes stop appearing in SSH sessions, my BBC News video feed pauses and - most annoyingly - when calls on my mobile go dead for seconds at a time (I am using a Vodafone femtocell).

The following illustrates the problem:

$ ping -c 50 www.bbc.co.uk
[...]
--- www.bbc.net.uk ping statistics ---
50 packets transmitted, 47 received, 6% packet loss, time 49095ms
rtt min/avg/max/mdev = 12.346/15.767/27.097/2.598 ms

$ ping -c 50 www.google.com
--- www-tmmdi.l.google.com ping statistics ---
50 packets transmitted, 40 received, 20% packet loss, time 49131ms
rtt min/avg/max/mdev = 31.685/34.465/40.997/2.109 ms

>pathping -n www.bbc.co.uk

Tracing route to www.bbc.net.uk [212.58.246.160]
over a maximum of 30 hops:
0 10.85.0.136
1 10.85.0.1
2 82.46.93.1
3 80.1.243.237
4 195.188.230.49
5 80.1.241.9
6 213.105.175.157
7 213.105.172.130
8 62.253.185.78
9 212.58.239.249
10 212.58.238.149
11 212.58.239.234
12 212.58.255.12
13 212.58.246.160

Computing statistics for 325 seconds...
Source to Here This Node/Link
Hop RTT Lost/Sent = Pct Lost/Sent = Pct Address
0 10.85.0.136
0/ 100 = 0% |
1 0ms 0/ 100 = 0% 0/ 100 = 0% 10.85.0.1
0/ 100 = 0% |
2 12ms 0/ 100 = 0% 0/ 100 = 0% 82.46.93.1
0/ 100 = 0% |
3 15ms 0/ 100 = 0% 0/ 100 = 0% 80.1.243.237
8/ 100 = 8% |
4 14ms 9/ 100 = 9% 1/ 100 = 1% 195.188.230.49
0/ 100 = 0% |
5 11ms 8/ 100 = 8% 0/ 100 = 0% 80.1.241.9
0/ 100 = 0% |
6 14ms 10/ 100 = 10% 2/ 100 = 2% 213.105.175.157
0/ 100 = 0% |
7 14ms 12/ 100 = 12% 4/ 100 = 4% 213.105.172.130
0/ 100 = 0% |
8 16ms 10/ 100 = 10% 2/ 100 = 2% 62.253.185.78
0/ 100 = 0% |
9 24ms 9/ 100 = 9% 1/ 100 = 1% 212.58.239.249
0/ 100 = 0% |
10 22ms 9/ 100 = 9% 1/ 100 = 1% 212.58.238.149
0/ 100 = 0% |
11 15ms 9/ 100 = 9% 1/ 100 = 1% 212.58.239.234
0/ 100 = 0% |
12 16ms 8/ 100 = 8% 0/ 100 = 0% 212.58.255.12
1/ 100 = 1% |
13 16ms 9/ 100 = 9% 0/ 100 = 0% 212.58.246.160

Trace complete.

Happy to provide more info if it helps to fix the problem.

Thanks

Sephiroth
19-05-2010, 14:59
Pathping without the -n would help assessment - we do need node names or at least I do!

Also your modem stats and event log from 192.168.100.1 (obscuring your CM MAC address if it appears). The packet loss that potehntiall matters on ICMP are those packets transmitted forward to the intended node. Packets discarded at the busy node don't matter.

sparksy
19-05-2010, 19:01
Downstream:

Downstream Lock : Locked
Downstream Channel Id : 8
Downstream Frequency : 323000000 Hz
Downstream Modulation : QAM256
Downstream Symbol Rate : 6952 Ksym/sec
Downstream Interleave Depth : taps12Increment17
Downstream Receive Power Level : 4.9 dBmV
Downstream SNR : 41.5 dB

Upstream:


Upstream Lock : Locked
Upstream Channel ID : 1
Upstream Frequency : 25800000 Hz
Upstream Modulation : QAM16
Upstream Symbol Rate : 2560 Ksym/sec
Upstream transmit Power Level : 40.5 dBmV
Upstream Mini-Slot Size : 2

Modem info:

Cable Modem : Euro-DOCSIS 1.0/1.1/2.0 Compliant
MAC Address :
Serial Number :
Boot Code Version : 1.1.2c
Software Version : 2.111.1002
Hardware Version : 1.9

Event log:

Wed May 19 10:11:57 2010 Wed May 19 10:11:57 2010 Information (7) The s/w filename specified in the config file is the same as ...
Wed May 19 10:11:57 2010 Wed May 19 10:11:57 2010 Information (7) A software upgrade filename was specified in the config file.
Wed May 19 10:11:57 2010 Wed May 19 10:11:57 2010 Information (7) Authorized
Wed May 19 10:11:57 2010 Wed May 19 10:11:57 2010 Information (7) Registration complete!
Wed May 19 10:11:57 2010 Wed May 19 10:11:57 2010 Information (7) We registered with a DOCSIS 1.1 config file!
Wed May 19 10:11:57 2010 Wed May 19 10:11:57 2010 Information (7) Received a REG-RSP message from the CMTS...
Wed May 19 10:11:56 2010 Wed May 19 10:11:56 2010 Information (7) Sending a REG-REQ to the CMTS...
Wed May 19 10:11:56 2010 Wed May 19 10:11:56 2010 Information (7) CableModem SNMP configure complete
Wed May 19 10:11:56 2010 Wed May 19 10:11:56 2010 Information (7) IP init completed ok
Wed May 19 10:11:56 2010 Wed May 19 10:11:56 2010 Information (7) CableModem TFTP init ok
Wed May 19 10:11:56 2010 Wed May 19 10:11:56 2010 Information (7) CableModem DHCP client init ok
Wed May 19 10:11:56 2010 Wed May 19 10:11:56 2010 Critical (3) DHCP WARNING - Non-critical field invalid in response.
Time Not Established Time Not Established Information (7) MAP w/initial maintenance region received
Time Not Established Time Not Established Information (7) Downstream sync ok
Time Not Established Time Not Established Information (7) Beginning initial ranging...
Time Not Established Time Not Established Information (7) downstream time sync acquired...
Time Not Established Time Not Established Information (7) Downstream sync ok
Time Not Established Time Not Established Information (7) starting ds time sync acquisition...
Time Not Established Time Not Established Information (7) Locked on the downstream. Waiting for UCDs...
Time Not Established Time Not Established Information (7) Downstream lock ok
Time Not Established Time Not Established Information (7) Sync Start
Wed May 19 10:09:58 2010 Wed May 19 10:09:58 2010 Critical (3) DHCP WARNING - Non-critical field invalid in response.
Wed May 12 14:12:17 2010 Wed May 12 14:12:17 2010 Critical (3) DHCP WARNING - Non-critical field invalid in response.
Thu Apr 29 07:39:59 2010 Thu Apr 29 07:39:59 2010 Critical (3) DHCP WARNING - Non-critical field invalid in response.
Wed Apr 28 23:46:55 2010 Wed Apr 28 23:46:55 2010 Critical (3) DHCP WARNING - Non-critical field invalid in response.
Wed Apr 28 23:44:50 2010 Wed Apr 28 23:44:50 2010 Critical (3) No Ranging Response received - T3 time-out
Wed Apr 28 23:44:45 2010 Wed Apr 28 23:44:45 2010 Critical (3) No Ranging Response received - T3 time-out
Wed Apr 28 23:44:44 2010 Wed Apr 28 23:44:44 2010 Critical (3) No Ranging Response received - T3 time-out
Wed Apr 28 23:44:42 2010 Wed Apr 28 23:44:42 2010 Critical (3) No Ranging Response received - T3 time-out
Wed Apr 28 23:44:36 2010 Wed Apr 28 23:44:36 2010 Critical (3) DHCP FAILED - Discover sent, no offer received
Wed Apr 28 23:41:26 2010 Wed Apr 28 23:41:26 2010 Critical (3) No Ranging Response received - T3 time-out
Wed Apr 28 23:44:38 2010 Wed Apr 28 23:44:38 2010 Critical (3) SYNC Timing Synchronization failure - Failed to acquire QAM/Q...

Pingpath:

>pathping www.bbc.co.uk

Tracing route to www.bbc.net.uk [212.58.246.158]
over a maximum of 30 hops:
0 foo.internal [10.85.0.136]
1 router [10.85.0.1]
2 cpc2-stav16-2-0-gw.aztw.cable.virginmedia.com [82.46.93.1]
3 osr01stav-ge112.network.virginmedia.net [80.1.243.109]
4 osr01azte-tenge71.network.virginmedia.net [195.188.230.49]
5 aztw-core-1a-ae0-0.network.virginmedia.net [80.1.241.9]
6 winn-bb-1a-as0-0.network.virginmedia.net [213.105.175.157]
7 glfd-bb-1b-so-100-0.network.virginmedia.net [213.105.172.130]
8 redb-ic-1-as0-0.network.virginmedia.net [62.253.185.78]
9 212.58.239.249
10 * 212.58.238.149
11 te12-1.hsw1.cwwtf.bbc.co.uk [212.58.239.234]
12 212.58.255.12
13 fmt-vip01.cwwtf.bbc.co.uk [212.58.246.158]

Computing statistics for 325 seconds...
Source to Here This Node/Link
Hop RTT Lost/Sent = Pct Lost/Sent = Pct Address
0 foo.internal [10.85.0.136]
0/ 100 = 0% |
1 0ms 0/ 100 = 0% 0/ 100 = 0% router [10.85.0.1]
0/ 100 = 0% |
2 13ms 0/ 100 = 0% 0/ 100 = 0% cpc2-stav16-2-0-gw.aztw.cable.virginmedia.com [82.46.93.1]
0/ 100 = 0% |
3 17ms 0/ 100 = 0% 0/ 100 = 0% osr01stav-ge112.network.virginmedia.net [80.1.243.109]
3/ 100 = 3% |
4 22ms 3/ 100 = 3% 0/ 100 = 0% osr01azte-tenge71.network.virginmedia.net [195.188.230.49]
0/ 100 = 0% |
5 12ms 3/ 100 = 3% 0/ 100 = 0% aztw-core-1a-ae0-0.network.virginmedia.net [80.1.241.9]
0/ 100 = 0% |
6 15ms 4/ 100 = 4% 1/ 100 = 1% winn-bb-1a-as0-0.network.virginmedia.net [213.105.175.157]
0/ 100 = 0% |
7 14ms 3/ 100 = 3% 0/ 100 = 0% glfd-bb-1b-so-100-0.network.virginmedia.net [213.105.172.130]
0/ 100 = 0% |
8 15ms 3/ 100 = 3% 0/ 100 = 0% redb-ic-1-as0-0.network.virginmedia.net [62.253.185.78]
0/ 100 = 0% |
9 25ms 3/ 100 = 3% 0/ 100 = 0% 212.58.239.249
0/ 100 = 0% |
10 20ms 4/ 100 = 4% 1/ 100 = 1% 212.58.238.149
0/ 100 = 0% |
11 16ms 4/ 100 = 4% 1/ 100 = 1% te12-1.hsw1.cwwtf.bbc.co.uk [212.58.239.234]
0/ 100 = 0% |
12 16ms 3/ 100 = 3% 0/ 100 = 0% 212.58.255.12
1/ 100 = 1% |
13 16ms 4/ 100 = 4% 0/ 100 = 0% fmt-vip01.cwwtf.bbc.co.uk [212.58.246.158]

Trace complete.

Ignitionnet
19-05-2010, 19:42
Issue:

3 15ms 0/ 100 = 0% 0/ 100 = 0% 80.1.243.237
8/ 100 = 8% |
4 14ms 9/ 100 = 9% 1/ 100 = 1% 195.188.230.49
0/ 100 = 0% |

Seems to be on link between Staverton and Bristol - Digital Fanatic can probably advise further.

The interfaces are:

osr01stav-tenge71.network.virginmedia.net

and

osr01azte-tenge71.network.virginmedia.net

Hopefully someone with a login to those two can test across that 10GbE interface and confirm / deny.

Also worth looking at is checking the routing between your IP range, 82.46.93.1, and osr01azte, IE either pinging osr01azte from your default gateway or vice versa, and checking osr01azte's routing table to ensure there's no routes flapping.

sparksy
19-05-2010, 21:18
Ping osr01azte from here:

--- osr01azte-tenge71.network.virginmedia.net ping statistics ---
50 packets transmitted, 49 received, 2% packet loss, time 49079ms
rtt min/avg/max/mdev = 8.289/15.045/136.002/18.715 ms

DarthMuppet
19-05-2010, 23:50
Thanks all. As luck (or misfortune depending on your point of view) today's service was perfect. Until 4.21pm when the whole thing fell to pieces in me. AS i had a need to get to the post office before it closed, i took that time to go, but whilst I was away I lost 60% of my ICMP in the first 30 mins and then it was almost perfect and still is. When it goes wrong, it does it in some style, and I'll pass back the answers to your suggestions as I am able to (assuming it goes wrong again).

Pathping to BBC
C:\Users\David>pathping -n bbc.co.uk

Tracing route to bbc.co.uk [212.58.224.138]
over a maximum of 30 hops:
0 213.81.81.41
1 213.81.80.1
2 80.1.201.93
3 212.43.163.141
4 213.105.172.13
5 62.253.184.2
6 212.58.239.237
7 212.58.238.153
8 212.58.224.138

Computing statistics for 200 seconds...
Source to Here This Node/Link
Hop RTT Lost/Sent = Pct Lost/Sent = Pct Address
0 213.81.81.41
6/ 100 = 6% |
1 144ms 6/ 100 = 6% 0/ 100 = 0% 213.81.80.1
0/ 100 = 0% |
2 137ms 6/ 100 = 6% 0/ 100 = 0% 80.1.201.93
0/ 100 = 0% |
3 177ms 6/ 100 = 6% 0/ 100 = 0% 212.43.163.141
0/ 100 = 0% |
4 139ms 7/ 100 = 7% 1/ 100 = 1% 213.105.172.13
0/ 100 = 0% |
5 174ms 6/ 100 = 6% 0/ 100 = 0% 62.253.184.2
0/ 100 = 0% |
6 194ms 6/ 100 = 6% 0/ 100 = 0% 212.58.239.237
1/ 100 = 1% |
7 178ms 7/ 100 = 7% 0/ 100 = 0% 212.58.238.153
0/ 100 = 0% |
8 161ms 7/ 100 = 7% 0/ 100 = 0% 212.58.224.138

Trace complete.

shows the route but no errors

C/M data
Downstream

Frequency 299000000 307000000 315000000 323000000
Lock Status
(QAM Lock/FEC Sync/MPEG Lock) Y/Y/Y Y/Y/Y Y/Y/Y Y/Y/Y
Channel Id 21 22 23 24
Modulation 256QAM 256QAM 256QAM 256QAM
Symbol Rate
(Msym/sec) 6.952 6.952 6.952 6.952
Interleave Depth I=12
J=17 I=12
J=17 I=12
J=17 I=12
J=17
Power Level
(dBmV) 9.67 9.56 8.84 9.23
RxMER
(dB) 37.36 37.09 37.94 37.94
Correctable
Codewords 3 12 1 2
Uncorrectable
Codewords 299 393 257 257

Upstream

US-1 US-2 US-3 US-4
Channel Type 1.0 N/A N/A N/A
Channel ID 4 N/A N/A N/A
Frequency
(Hz) 47400000 N/A N/A N/A
Ranging Status Success N/A N/A N/A
Modulation 16QAM N/A N/A N/A
Symbol Rate
(KSym/sec) 2560 N/A N/A N/A
Mini-Slot Size 2 N/A N/A N/A
Power Level
(dBmV) 54.00 N/A N/A N/A
T1 Timeouts 0
T2 Timeouts 0 0 0 0
T3 Timeouts 0 0 0 0
T4 Timeouts 0 0 0 0

Modem info

Cable Modem : DOCSIS 1.0/1.1/2.0/3.0 Compliant
Serial Number :
Boot Code Version : 6.1.1f
Software Version : 3.11.1011
Hardware Version : 1.39
CA Key : Installed

Event log. These seem to come up in pairs a few seconds apart.

Tue May 18 15:39:18 2010 Tue May 18 15:39:18 2010 Warning (5) TEK Invalid - Invalid Key Sequence Number;CM-MAC=00:22:68:xx:xx:xx;CMTS-MAC=00:30:b8:d1:f6:90;CM-QOS=1.1;CM-VER=3.0;
Tue May 18 15:39:28 2010 Tue May 18 15:39:28 2010 Critical (3) No Ranging Response received - T3 time-out;CM-MAC=00:22:68:xx:xx:xx;CMTS-MAC=00:30:b8:d1:f6:90;CM-QOS=1.1;CM-VER=3.0;

a quick google suggests these might be relevant, but they don't see to match the times when i get problems.

Tracerts


Tracing route to osr01azte-tenge71.network.virginmedia.net [195.188.230.49]
over a maximum of 30 hops:

1 102 ms 1 ms 3 ms DC-WRT [192.168.0.1]
2 11 ms 13 ms 29 ms cpc2-papw5-2-0-gw.know.cable.virginmedia.com [213.81.80.1]
3 13 ms 11 ms 15 ms cmbg-core-1b-ge-210-2252.network.virginmedia.net [80.1.201.221]
4 13 ms 16 ms 11 ms nrth-bb-1b-ae5-0.network.virginmedia.net [212.43.163.145]
5 26 ms 20 ms 17 ms popl-bb-1a-as3-0.network.virginmedia.net [213.105.172.14]
6 27 ms 29 ms 16 ms popl-bb-1b-ae0-0.network.virginmedia.net [213.105.174.230]
7 26 ms 23 ms 20 ms winn-bb-1a-as1-0.network.virginmedia.net [212.43.162.193]
8 27 ms 21 ms 28 ms aztw-core-1a-as0-0.network.virginmedia.net [213.105.175.158]
9 17 ms 27 ms 22 ms osr01azte-tenge71.network.virginmedia.net [195.188.230.49]

Trace complete.


and

Tracing route to osr01stav-tenge71.network.virginmedia.net [195.188.230.50]
over a maximum of 30 hops:

1 162 ms 1 ms 1 ms DC-WRT [192.168.0.1]
2 10 ms 20 ms 8 ms cpc2-papw5-2-0-gw.know.cable.virginmedia.com [213.81.80.1]
3 15 ms 15 ms 17 ms cmbg-core-1a-ge-210-2251.network.virginmedia.net [80.1.201.93]
4 14 ms 19 ms 22 ms popl-bb-1a-ae2-0.network.virginmedia.net [212.43.163.141]
5 53 ms 20 ms 19 ms popl-bb-1b-ae0-0.network.virginmedia.net [213.105.174.230]
6 25 ms 23 ms 17 ms winn-bb-1a-as1-0.network.virginmedia.net [212.43.162.193]
7 19 ms 18 ms 24 ms aztw-core-1a-as0-0.network.virginmedia.net [213.105.175.158]
8 26 ms 18 ms 18 ms osr01azte-pc111.network.virginmedia.net [80.1.241.10]
9 21 ms 51 ms 28 ms osr01stav-tenge71.network.virginmedia.net [195.188.230.50]

Trace complete.


Observation
first hop is 30msec. in my corporate world of high-speed Cisco-based ethernet and international MPLS links, I can get from my desk in the UK to the east coast of the USA in 60 msec. that's 30 times the distance in just twice the time. is that long latency to the first hop inherent in cable networks? Is it VM's design? is it a fault? I have no clue.

MTU
MTU is 1300 cos Cisco's VPN client decides it must be. in practice the setting makes no observable difference to the fault. I may set it back to 1500, as i like efficiency, and don't use the VPN client too much. I can't split the tunnel, so it's a bit restrictive.