Good day everyone
We have situation where one of our vfiler has many exports to different subnet with huge capacity of volumes/qtrees but mostly for unix
one such qtree has been exported using a different name with the actual path like below
abcd -actual=/vol/myexport/abcd,sec=sys,rw=10.0.0.0/24:10.0.1.0/24,root=10.0.0.0/24
it was mounted 100+ Suse clients
once there was takeover/giveback in the filer side ....all mounts got hung and somehow it was unmounted from the os side ..and couples of them used automount so it was already unmount situation
now problem is
whereever this "acbd" was mounted before ...remounting can not be done with TCP and nfsV3..it gets hung at UDP port 4046 and comes out after sometimes "timeout" error
some clinet it can be remounted using V2 with tcp
and if we use UDP ..then it can be remounted anywhere
so 100+ clients are affected with same issue
but it is a new client where it was not mounted before TO/GB ..it works fine .with NFSV3 ...
does anyone has any clue on this ? or face same issue ?
note :
on the same clinet ...there are multiple exports are mounted from same vfilers and working fine
7mode -8.2.3P2
metro cluster
