Ping Check Problems - Lost Packets?

[tomrhyne]

I frequently have long delays before a web page downloads. I have installed both recommended patches from this site. Speed is ok (I have 500kbs cable modem). When I run Tweek Test on DSLReports I frequently get results like this:

ICMP (ping) check Minimum ping: 52 ms
Maximum ping: 58 ms
Ping stability:57 58 55 52 - - - - - -
Loss at 56byte pkts: 6%
Loss at 1000byte pkts: 30%

The rest of the results are:


Your Tweakable Settings:

Receive Window (RWIN): 531440
Window Scaling: 4
Path MTU Discovery: ON
RFC1323 Window Scaling: ON
RFC1323 Time Stamping: OFF
Selective Acks: ON
MSS requested: 1460
TTL: 52(less any hops behind firewall)
TTL remaining: 37


Example 146000 byte download

Actual data bytes sent: 157680
Actual data packets: 108
Max packet sent (MTU): 1500
Max packet recd (MTU): 1500
Retransmitted data packets: 8
sacks you sent: 23
pushed data pkts: 1
data transmit time: 2.435 secs
our max idletime: 110.3 ms
transfer rate: 52135 bytes/sec
transfer rate: 417 kbits/sec(This is not a speed test!transfer efficiency: 92%)

Are the loss of packets due to my ISP? What action should I take? I don't have an alternative ISP.

[DaveM]

your need to run a continuous traceroute program that will report packet loss. The first hop with any significant packet loss is general very close to the problem.

All hops after the first affected hop, will also have packet loss, due to the first hop losing responses from any subsequent latter server!

[MosDef112]

You need to lower your RWIN to 64240. You're experiencing packet loss because your receive buffer is too large for your Internet line speed.

[tomrhyne]

Thank you for your response. I will try the lower RWIN.

[tomrhyne]

Tried lowering RWIN as suggested, but smae results.

ICMP (ping) check
Minimum ping: 53 ms
Maximum ping: 74 ms
Ping stability:
54 56 59 53 53 54 74 - - -
Loss at 56byte pkts: 17%
Loss at 1000byte pkts: 30%



Your Tweakable Settings:
Receive Window (RWIN): 64240
Window Scaling: 0
Path MTU Discovery: ON
RFC1323 Window Scaling: ON
RFC1323 Time Stamping: OFF
Selective Acks: ON
MSS requested: 1460
TTL: 52
(less any hops behind firewall)
TTL remaining: 37



Example 146000 byte download
Actual data bytes sent: 157680
Actual data packets: 108
Max packet sent (MTU): 1500
Max packet recd (MTU): 1500
Retransmitted data packets: 8
sacks you sent: 22
pushed data pkts: 5
data transmit time: 3.260 secs
our max idletime: 321.9 ms
transfer rate: 39736 bytes/sec
transfer rate: 317 kbits/sec
This is not a speed test!
transfer efficiency: 92%

[MosDef112]

Is your service PPPoE or static IP?

[tomrhyne]

It is nor PPPoE. So I assume it must be static IP. ISP is Plantationcable.net provided by HSA Corp out of Louiseville, KY. Domain server is AUTH0.NS.HSACORP.NET & AUTH1.NS.HSACORP.NET.

[tomrhyne]

Changed RWIN to 13140 and still have problems.


ICMP (ping) check
Minimum ping: 52 ms
Maximum ping: 58 ms
Ping stability:
52 54 57 58 52 53 - - - -
Loss at 56byte pkts: 15%
Loss at 1000byte pkts: 30%


Your Tweakable Settings:
Receive Window (RWIN): 13140
Window Scaling: off
Path MTU Discovery: ON
RFC1323 Window Scaling: OFF
RFC1323 Time Stamping: OFF
Selective Acks: ON
MSS requested: 1460
TTL: 56
(less any hops behind firewall)
TTL remaining: 37


RWIN is in range
Looking good


Example 146000 byte download
Actual data bytes sent: 150380
Actual data packets: 103
Max packet sent (MTU): 1500
Max packet recd (MTU): 1500
Retransmitted data packets: 3
sacks you sent: 16
pushed data pkts: 20
data transmit time: 2.654 secs
our max idletime: 134.4 ms
transfer rate: 48061 bytes/sec
transfer rate: 384 kbits/sec
This is not a speed test!
transfer efficiency: 97%


A few re-transmissions seen (FAQ #705)

[MosDef112]

Run a few traces in here.. Open a MS-DOS prompt, and type the following:

C:\WINDOWS\TRACERT OMARNYC.COM > c:\trace1.txt
C:\WINDOWS\TRACERT PCMAGLABS.COM > c:\trace2.txt

Copy and paste trace1.txt and trace2.txt in here, change your first hop so no one knows your IP address.

[Kip Patterson]

What MosDef112 is shooting for is looking from your end to see where the packet loss occurs. That is a lot more meaningful than a check from the far end.

I did a traceroute to you and saw no packet loss, but there was a 300ms latency in Atlanta.

Kip

[tomrhyne]

I tried doing as you said, but with \trace1.txt the trace would run but no data would come back. I eliminated trace1.txt and ran the command. I could never get a complete trace because the trace would repaetedly time out after about the 13th hop somewhere after a ny server. I tried the trace to yahoo with the following results:

Tracing route to yahoo.com [216.115.108.245]
over a maximum of 30 hops:

1 <10 ms 14 ms 13 ms 24-216-1-1.hsacorp.net [24.216.1.1]
2 27 ms 27 ms 28 ms 519.hssi2-0-0.GW1.ATL1.alter.net [157.130.8
]
3 27 ms 28 ms 27 ms 174.at-1-0-0.XR2.ATL1.ALTER.NET [152.63.82.
4 41 ms 14 ms 28 ms 194.at-2-0-0.XR2.ATL5.ALTER.NET [152.63.80.

5 27 ms 41 ms 14 ms 192.ATM6-0.BR3.ATL5.ALTER.NET [152.63.82.21
6 41 ms 41 ms 41 ms atm2-0-1.core1.Atlanta1.Level3.net [209.0.2
1]
7 55 ms 27 ms 55 ms lo0.mp1.Atlanta1.Level3.net [209.247.8.247]
8 96 ms 96 ms 96 ms so-0-0-0.mp2.SanJose1.Level3.net [209.247.8

9 83 ms 110 ms 96 ms gigabitethernet10-0.ipcolo4.SanJose1.Level3
[64.159.2.42]
10 * * * Request timed out.
11 96 ms 96 ms 96 ms ge-2-3-0.msr2.pao.yahoo.com [216.115.101.46
12 96 ms 96 ms 110 ms vlan28.bas1-m.snv.yahoo.com [216.115.100.12
13 96 ms 110 ms 96 ms img5.yahoo.com [216.115.108.245]

[tomrhyne]

This is my trace to PCMAGLABS.COM. Notice a couple long delays. I did it once more and got no long delays. Still cannot get to OMARNYC.COM.


2 27 ms 28 ms 27 ms 552.Hssi2-0-0.GW1.ATL3.ALTER.NET [157.130.80.16
]
3 13 ms 28 ms 14 ms 147.at-1-1-0.XR2.ATL1.ALTER.NET [152.63.81.62]
4 27 ms 28 ms 27 ms 0.so-3-0-0.TR2.ATL1.ALTER.NET [152.63.10.77]
5 41 ms 41 ms 27 ms 109.at-5-0-0.TR2.CHI2.ALTER.NET [146.188.141.46

6 41 ms 55 ms 41 ms 0.so-3-0-0.XR2.CHI2.ALTER.NET [152.63.15.133]
7 41 ms 28 ms 68 ms 194.ATM7-0.BR2.CHI1.ALTER.NET [152.63.65.173]
8 55 ms 41 ms 55 ms 207.88.50.57
9 55 ms 27 ms 41 ms ge9-0.tran1.chi-il.us.xo.net [64.220.0.177]
10 41 ms 41 ms 42 ms ge10-0.tran2.chi-il.us.xo.net [64.220.0.178]
11 55 ms 69 ms 68 ms p1-0.tran2.nwk-nj.us.xo.net [64.0.0.46]
12 385 ms 69 ms 54 ms p2-0.dist1.hud-ny.us.xo.net [64.0.0.62]
13 110 ms 96 ms 69 ms ge12-0-0.access1.hud-ny.us.xo.net [64.220.3.86]

14 68 ms 96 ms 83 ms rb-209-220-116-42.hud-ny.us.xo.net [209.220.116
42]
15 83 ms 82 ms 96 ms w155.z209220187.nyc-ny.dsl.cnc.net [209.220.187
155]
16 82 ms 82 ms 96 ms w170.z064000229.nyc-ny.dsl.cnc.net [64.0.229.17
]

[MosDef112]

I just noticed that OMARNYC.COM is down, sorry about that. Your path from your end to PCMAGLABS.COM seems okay. It does take an unusual hop from Atlanta, all the way to Chicago with no additional hops in between. So perhaps it may be a very long fiber line that sometimes becomes congested. Try tracing this other address, 64.81.221.182, and post results here.

[tomrhyne]

Tracing route to dsl081-221-182.nyc2.dsl.speakeasy.net [64.81.221.182]
over a maximum of 30 hops:

1 27 ms 14 ms 27 ms 24-216-1-1.hsacorp.net [24.216.1.1]
2 28 ms 13 ms 28 ms 519.hssi2-0-0.GW1.ATL1.alter.net [157.130.84.22
]
3 27 ms 14 ms 27 ms 174.at-1-0-0.XR2.ATL1.ALTER.NET [152.63.82.54]
4 27 ms 28 ms 14 ms 0.so-3-0-0.TR2.ATL1.ALTER.NET [152.63.10.77]
5 55 ms 27 ms 42 ms 109.at-5-0-0.TR2.NYC9.ALTER.NET [146.188.141.78

6 55 ms 55 ms 41 ms 286.at-2-0-0.XR2.BOS4.ALTER.NET [152.63.19.117]

7 55 ms 42 ms 54 ms 178.ATM7-0.GW5.BOS4.ALTER.NET [152.63.25.61]
8 55 ms 68 ms 42 ms internap-gw.customer.alter.net [65.194.79.74]
9 82 ms 69 ms 55 ms border4.ge3-0-bbnet2.bsn.pnap.net [63.251.128.7
]
10 68 ms 69 ms 69 ms spk-1-bos.dsl-isp.net [63.251.141.193]
11 97 ms 68 ms 83 ms dsl081-204-151.nyc2.dsl.speakeasy.net [64.81.20
.151]
12 83 ms 68 ms 83 ms dsl081-221-182.nyc2.dsl.speakeasy.net [64.81.22
.182]

[MosDef112]

From what I see, all your traces seem fine. I would think it's a DNS server issue. Try using different DNS servers. If you can, hard code the following addresses:

216.168.224.200
216.168.254.69