Re: [SSI-devel] Interesting scenario in case of Failover with DRBD
Brought to you by:
brucewalker,
rogertsang
From: Aneesh K. <ane...@gm...> - 2005-06-01 12:35:18
|
On 6/1/05, Roopa Prabhu <pra...@in...> wrote: > > > STONITH ( Shoot the other node in the head ) :). BTW in OpenSSI you > > can drop node mionitoring frequency. Roopa should have details. >=20 > echo 1 > /proc/cluster/nm_nodedown_disabled > Instead i guess is nice to change the node monitor rate. I take two values. I am lazy to dig up which way. But one is how many sec between I am alive packet and the second how many retries before declaring node down. -aneesh |