Re: super slow speeds from director

To: " users mailing list." <lvs-users@xxxxxxxxxxxxxxxxxxxxxx>
Subject: Re: super slow speeds from director
From: Rob <ipvsuser@xxxxxxxxxxxxxxxx>
Date: Thu, 22 Feb 2007 18:21:35 -0800
Roberto Nibali wrote:
No handshake visible, please capture the whole TCP flow.

from home, www: (normal speed)

from home, wwwdb1: (normal speed)

All dumps were done on wwwdb1 when accessing. Hopefully you guys can make heads/tails out of those dumps cause I don't know what I'm reading.

Hmm, could you add at least one '-v'? To me it's not enough information to figure out what's going.

I did notice that in both cases, going directly to wwwdb1 used about 200 packets and both cases of using the cluster, around 100.

On the cluster don't you want to capture from the ethernet interface level on the director to really see what is being returned to the client? A dump from the RS might not look that different between the two cases but who knows what is going on on the director.

tcpdump -s 160 -X -nnvvvi eth? host clientIP and port 80

to get both sides of the connection and the -s/-X will show some of the query and response so it is easier to line up what part of the tcp dump corresponds to what part of the http flow and if there are major differences between the VIP and RIP sequences.


<Prev in Thread] Current Thread [Next in Thread>