Uh, whats going on here? ICMP replies negative times.

[EULA]

Recently I noticed that web pages were coming up rather slowly, and after a virus and spyware scan turned up nothing, I started to think the problem might be hardware related, so I did a few tests and recieved these interesting results:

Ping to Router

Pinging 10.0.0.1 with 32 bytes of data:

Reply from 10.0.0.1: bytes=32 time<1ms TTL=64
Reply from 10.0.0.1: bytes=32 time=-96ms TTL=64
Reply from 10.0.0.1: bytes=32 time<1ms TTL=64
Reply from 10.0.0.1: bytes=32 time=-96ms TTL=64

Ping statistics for 10.0.0.1:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 0ms, Maximum = -96ms, Average = 1073741776ms

Loopback

Pinging 127.0.0.1 with 32 bytes of data:

Reply from 127.0.0.1: bytes=32 time=97ms TTL=128
Reply from 127.0.0.1: bytes=32 time=-96ms TTL=128
Reply from 127.0.0.1: bytes=32 time<1ms TTL=128
Reply from 127.0.0.1: bytes=32 time=97ms TTL=128

Ping statistics for 127.0.0.1:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 0ms, Maximum = -96ms, Average = 24ms

I figured it was probably just a flakey card (Built-in nForce) so I disabled that adapter and inserted an Intel Pro 100+ PCI card and the problem seemed to be fixed for about a day. Again today, I noticed webpages were slow and unresponsive, so I tried the above two tests, and it's the same problem.

Am I just so lucky enough to have two bad NICs, or is there something that I'm missing? I've tried uninstalling/reinstalling/updating drivers to no avail. I think I have a 3Com card around somewhere that I can try.

The only other thing I can think to try is to rebuild the TCP/IP stack which I am going to attempt now.

 

[EULA]

Alright, I rebuild the TCP/IP stack, but that didn't change anything, so I rebuilt the Winsock LSP chain using netsh Winsock reset, and so far everything seems to be pretty peachy at time<1ms.


I'll followup later and see if this lasts.

 

[EULA]

I awoke this morning to find that the times had reached -20ms...

 

[Genx87]

I had an ASUS Nforce 4 MB that exhibits this behavior with the Nforce ethernet chipset. The other built in ethernet adapter doesnt.

The problem i found with the Nforce chipset when it produced they negative pings is it would hang network browsing and UNC paths.

keep us updated on what you find. I simple disabled the thing and use the other card.

 

[EULA]

I'm going to try a third card, in the event the intel card I'm using now is bad (It's doing the same thing) and will return.

 

[cmetz]

EULA, that's a software bug.

 

[EULA]

What software would it be then? I've rebuilt TCP/IP and Winsock, but the issue returns. I've already tried testing with my software firewall disabled, but I wouldnt think that would affect a loopback test anyway.

Its not like its just one or replies that does this, its about 75%. I can run the same test on another workstation or server, and don't have the problem.

 

[Madwand1]

Try /usepmtimer in the boot.ini

http://support.microsoft.com/kb/895980

 

[EULA]

Originally posted by: Madwand1
Try /usepmtimer in the boot.ini

http://support.microsoft.com/kb/895980

It does make sense that it has something to do with a counter, because the value slowly increases (or decreases ). I'm running the test again right now.

 

[EULA]

It looks like that has done the trick; I don't know why it just started recently, nor why I haven't had this problem in the past. Thanks much!