Thread (128 messages) 128 messages, 16 authors, 2006-08-25

Re: [RFC][PATCH 0/9] Network receive deadlock prevention for NBD

From: Rik van Riel <hidden>
Date: 2006-08-13 01:11:37
Also in: linux-mm, lkml

David Miller wrote:
From: Peter Zijlstra <redacted>
Date: Sat, 12 Aug 2006 12:18:07 +0200
quoted
65535 sockets * 128 packets * 16384 bytes/packet = 
1^16 * 1^7 * 1^14 = 1^(16+7+14) = 1^37 = 128G of memory per IP

And systems with a lot of IP numbers are not unthinkable.
TCP restricts the amount of global memory that may be consumed
by all TCP sockets via the tcp_mem[] sysctl.
This is exactly why we need to be careful which sockets
we allocate memory for, when the system is about to run
out of memory.

-- 
"Debugging is twice as hard as writing the code in the first place.
Therefore, if you write the code as cleverly as possible, you are,
by definition, not smart enough to debug it." - Brian W. Kernighan
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help