[f-nsp] MLX32 To Extreme Transfer Issues

Wed Apr 8 15:03:28 EDT 2015

I am encountering an issue with a 40G trunk (4x10G dynamic LAG) between
MLX32 and a ExtremeXOS version 12.2.2.11 device.

The network on the Extreme side shows no issues locally. But, I see strange
behavior when traffic comes across the trunk to devices on the MLX32 side
of the network. UDP transfer of files via tftp is sluggish and sometimes
hangs. TCP transfer of files via rsync shows poor performance and sometimes
hangs. The issue will come and go and will only affect certain devices, not
all, at a given time.

Two ports from two distinct cards on the MLX32 connect to one card on the
Extreme. The issue does not isolate to a particular port and there is
plenty of capacity. The card on the Extreme switch was recently replaced.

Has anyone encountered something similar? Thanks in advance for your help!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://puck.nether.net/pipermail/foundry-nsp/attachments/20150408/00a1ee85/attachment.html>