[oe] Heavy packet loss using Slugos head

Mark Howells Mark.Howells at softoption.com
Sun Sep 17 09:38:35 UTC 2006


Hi,
 
I think my slug is dropping packets. I have been having lots of problems with NFS (slow performance, crashing). Messing with rsize,wsize settings for the mount seemed to point to larger packets being less reliable/slower. Flooging the slug (from a desktop Gentoo client) with a couple of ping statements gives the following
 
zeus ~ # ping -f -s 1400 192.168.0.5
PING 192.168.0.5 (192.168.0.5) 1400(1428) bytes of data.
....................................
--- 192.168.0.5 ping statistics ---
2327 packets transmitted, 2291 received, 1% packet loss, time 2543ms
rtt min/avg/max/mdev = 0.644/0.792/1.173/0.089 ms, ipg/ewma 1.093/0.739 ms

zeus ~ # ping -f -s 50000 192.168.0.5
PING 192.168.0.5 (192.168.0.5) 50000(50028) bytes of data.
......................................................................................
--- 192.168.0.5 ping statistics ---
204 packets transmitted, 118 received, 42% packet loss, time 2386ms
rtt min/avg/max/mdev = 10.863/11.218/11.643/0.167 ms, pipe 2, ipg/ewma 11.755/11.282 ms

Notice that the larger packet size is has a huge loss. Flooding another PC from the same client gives zero percent loss - so it looks like the slug is responsible for the loss. I'd appreciate a steer in the right direction to track this down as network analysis isn't my area of expertise.

Thanks

Mark

-- 
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.1.406 / Virus Database: 268.12.4/449 - Release Date: 15/09/2006
 




More information about the Openembedded-devel mailing list