Here are the times I have seen:
Packet Size RoundTrip Time
64 28
128 34
256 51
512 78
1024 142
1025 No response
Does this indicate a problem? That is, should the ping still work with
packet sizes in excess of 1024? If it ought to work, then why is it not
working in this instance?
Mitchell,
The max packet size for ping is 1024. What's happening is either your
ping program won't send packets over 1024bytes, or the machine you're
sending them to won't reply to them. This limitation got a lot of
publicity in the last year when it was discovered many operating
systems would react adversly (ie, hang, corrupt memory, etc.) when
receiving oversized ping packets (which became known as the "ping of
death"). Since then, most OSes and network hardware have been upgraded
to protect against this, but many of them will choose not to reply at
all instead.
Matthew
Sent via Deja.com http://www.deja.com/
Before you buy.
I was wondering if there was a maximum size. However, the reason I tried
this experiment in the first place is that a Techie at Bell Atlantic DSL told
me that he had tested my connection by pinging my gateway AND my IP address
with a packet size of 2004 bytes -- and he got responses from both computers.
(He did observe that the round-trip time to my IP address was substantially
slower than the time to the gateway.) That's why I tried pinging the gateway
from my "side" of the network -- and yet I can't do it large packets.
After I posted my initial message, I tried a different Ping utility (TJPing)
and with that I was able to ping the gateway using larger packets (up to
around 1400 bytes) but even with TJPing I ran into a limit on the size
packets I could use. (I have been able to ping various other Internet sites
with large packets with no problem -- it's just my DSL gateway that displays
this problem.
MW
---
Posted from dslreports.com... PC security scan, DSL info, and more.
Then increase the packet size until you get a fragmentation error. Ping
should still work with fragmented packets but the performance will be
poor.
You should find a MTU of 1500 bytes on BA. Sending larger packets than
the MTU on the path causes very poor performance.
--
Kent W. England
--
For a step-by-step guide to installing your own small
office network and shared Internet access using TCP/IP,
see www.6SigmaNets.com.
Mitchell Weitz <mitchw@bellatlanticdotnet> wrote in message
news:8E7151250mitch...@151.202.16.21...
> I have been testing the route to my default gateway (on Bell
Atlantic's
> infamous DSL service). I have found that I can ping my default
gateway
> successfully with packet sizes up to and including 1024 -- but if I
bump up
> the size to 1025 or greater, the test fails and I get total packet
loss.
>
>To find the MTU on your connection, set the ping option for "don't
>fragment". In Win9x this is "-f", which in Unix means "flood" so don't
>use that flag in Unix.
>
>Then increase the packet size until you get a fragmentation error. Ping
>should still work with fragmented packets but the performance will be
>poor.
You may find that any packet larger than 1500 (payload of 1460 in Win9x)
won't be returned.
>You should find a MTU of 1500 bytes on BA. Sending larger packets than
>the MTU on the path causes very poor performance.
>
>--
>Kent W. England
--
Best regards,
John Navas <http://navasgrp.home.att.net/>
CABLE/DSL TIPS: <http://navasgrp.home.att.net/tech/cable_dsl.htm>
--
Kent W. England
--
For a step-by-step guide to installing your own small
office network and shared Internet access using TCP/IP,
see www.6SigmaNets.com.
John Navas <spamf...@navasgrp.dublin.ca.us> wrote in message
news:1XYeOBHV=MiTWoUFOsG+D72NIN=a...@4ax.com...
I'm saying that many devices have "ping of death" filtering that discards
large packets.
"Kent W. England" <w98...@6SigmaNets.com> wrote:
>Because some OSs have ICMP bugs? If you try to ping with a packet size
>greater than your interface MTU, your own system will fragment the
>packet. Win9x does this just fine. Then the receiver must re-assemble
>this ping packet and send a response. You are saying that some OSs can't
>reassemble fragmented pings and respond? AFAIK, Win9x can re-assemble
>and respond.
>John Navas <spamf...@navasgrp.dublin.ca.us> wrote in message
>news:1XYeOBHV=MiTWoUFOsG+D72NIN=a...@4ax.com...
>> [POSTED TO comp.dcom.xdsl]
>> "Kent W. England" <w98...@6SigmaNets.com> wrote:
>>
>> >To find the MTU on your connection, set the ping option for "don't
>> >fragment". In Win9x this is "-f", which in Unix means "flood" so
>don't
>> >use that flag in Unix.
>> >
>> >Then increase the packet size until you get a fragmentation error.
>Ping
>> >should still work with fragmented packets but the performance will be
>> >poor.
>>
>> You may find that any packet larger than 1500 (payload of 1460 in
>Win9x)
>> won't be returned.
--
> I have been testing the route to my default gateway (on Bell Atlantic's
> infamous DSL service). I have found that I can ping my default gateway
> successfully with packet sizes up to and including 1024 -- but if I bump up
> the size to 1025 or greater, the test fails and I get total packet loss.
>
> Here are the times I have seen:
>
> Packet Size RoundTrip Time
> 64 28
> 128 34
> 256 51
> 512 78
> 1024 142
> 1025 No response
>
> Does this indicate a problem? That is, should the ping still work with
> packet sizes in excess of 1024? If it ought to work, then why is it not
> working in this instance?
Depends upon lots of factors. I've seen it stop at 8192 before. I've also
seen similar problems due to not having a firewall rule to allow fragmented
packets - but then you'd have real problems with your net as well.
Most likely nothing to worry about.
Gary
Steve
ping -s 40960 gateway.boulder.ibm.com
PING gateway.boulder.ibm.com (198.17.57.70): 40960 data bytes
40968 bytes from 198.17.57.70: icmp_seq=0 ttl=236 time=511.2 ms
40968 bytes from 198.17.57.70: icmp_seq=1 ttl=236 time=513.9 ms
40968 bytes from 198.17.57.70: icmp_seq=2 ttl=236 time=576.1 ms
40968 bytes from 198.17.57.70: icmp_seq=3 ttl=236 time=908.1 ms
40968 bytes from 198.17.57.70: icmp_seq=4 ttl=236 time=398.7 ms
40968 bytes from 198.17.57.70: icmp_seq=5 ttl=236 time=713.3 ms
40968 bytes from 198.17.57.70: icmp_seq=6 ttl=236 time=831.6 ms
40968 bytes from 198.17.57.70: icmp_seq=7 ttl=236 time=396.8 ms
40968 bytes from 198.17.57.70: icmp_seq=8 ttl=236 time=330.1 ms
40968 bytes from 198.17.57.70: icmp_seq=9 ttl=236 time=502.5 ms
40968 bytes from 198.17.57.70: icmp_seq=10 ttl=236 time=408.2 ms
--- gateway.boulder.ibm.com ping statistics ---
11 packets transmitted, 11 packets received, 0% packet loss
round-trip min/avg/max = 330.1/553.6/908.1 ms
________________________________________________________________________________
Anger is said to be a form of temporary insanity.
If you have the Time, and the inclination, do please visit me at
http://cpcug.org/user/ramster