]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/commit
NFS over TCP retransmit behavior should default to a 60 second time out,
authorChuck Lever <cel@FreeBSD.org>
Tue, 23 May 2006 18:48:07 +0000 (18:48 +0000)
committerChuck Lever <cel@FreeBSD.org>
Tue, 23 May 2006 18:48:07 +0000 (18:48 +0000)
commit6d0699a5ba8d92e0826eb71f98881d6c849db1b3
tree68e06f45da996223cc7b1442448c81d3941b537b
parent94163ea283bac49154a617d91d9a2849742a889c
NFS over TCP retransmit behavior should default to a 60 second time out,
mimicing the NFS reference implementation.

NFS over TCP does not need fast retransmit timeouts, since network loss
and congestion are managed by the transport (TCP), unlike with NFS over
UDP.  A long timeout prevents the unnecessary retransmission of non-
idempotent NFS requests.

Reviewed by: mohans, silby, rees?
Sponsored by: Network Appliance, Incorporated
sys/nfs4client/nfs4_vfsops.c
sys/nfsclient/nfs.h
sys/nfsclient/nfs_vfsops.c