Thread (182 messages) 182 messages, 27 authors, 2008-08-01

Re: [TCP bug, regression] stuck distcc connections in latest -git

From: David Miller <davem@davemloft.net>
Date: 2008-07-24 09:30:03
Also in: lkml

From: Ingo Molnar <redacted>
Date: Thu, 24 Jul 2008 11:25:18 +0200
ok, i've booted the latest kernel on dione and the builds are running on 
it now. It's using these settings currently:

 # cat /proc/sys/net/ipv4/tcp_syncookies
 0
 # cat /proc/sys/net/ipv4/tcp_sack
 1

Just to make sure i have the right test-plan: assuming this 
suspected-good kernel is trouble-free for say 48 hours, i should be able 
to switch back to the suspected-broken variant, by doing:

   echo 1 > /proc/sys/net/ipv4/tcp_syncookies

without rebooting the kernel, right? And then if the bug comes back 
again we've nailed it down to the area you suspect, correct?
You don't need the explicit tcp_sack=1, it's going to be '1' by
default.

Simply switching syncookies on and off will turn both SACK and
timestamps off when running the dist kernel.
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help