[cgl_discussion] Another POC for CFH.2.0 NFS Server Failover

[Zhu, Yi]

Hi,

Below link in linux-ha-dev list talked about support NFS Server Failover by linux-ha.
http://lists.community.tummy.com/pipermail/linux-ha-dev/2003-June/006000.html

The main issue to support this feature however, is not the cluster framework, it is
the nfs-utils version. statd in nfs-utils version earlier than 0.3.3 doesn't support the 
"-n" switch, which is used to specify the hostname the clients are mounting from.
Modern Linux releases are all sufficient with the requirement. Redhat 7.3 uses 
nfs-utils 0.3.3, UL 1.0 uses 1.0.1, and so on.

I have successfully setup the NFS Server Failover with linux-ha with a little 
modification with the configuration files. For detail information, please refer
http://lists.community.tummy.com/pipermail/linux-ha-dev/2003-June/006012.html

I did some basic test on the NFS Failover service. The service can be resumed on
the backup server. That is, all nfs mounted directory remains mounted after the 
failover, all the read/write request are queued during the takeover and are processed
after the takeover.

I did not test the authentication things because I'm not familar with that. Would
anyone tell me what is a NFS authenticated client?

I tried both kimberlite and linux-ha for the NFS Server Failover requirement. IMHO,
for the NFS Server Failover requirement, linux-ha can do as well as, if not better than
kimberlite. Shall we put linux-ha as another PoC for the CFH.2.0 NFS Server Failover
requirement?

Thanks,
-- 
------------------------------------------------------------------------------
Opinions expressed are those of the author and do not represent Intel Corp.

Zhu Yi (Chuyee)
Intel China Software Lab (ICSL)
22nd Floor, ShanghaiMart Tower No. 2299 Yan'an Road(West) Shanghai 200336, PRC
Tel: 8621-52574545-1261 Fax: 8621-62366119

GnuPG v1.0.6 (GNU/Linux)
http://cn.geocities.com/chewie_chuyee/gpg.txt
1024D/71C34820 C939 2B0B FBCE 1D51 109A  55E5 8650 DB90 71C3 4820