I have an HA cluster running into some hangs. It apears that it is haning
someplace in nfsauth portion. I have two questions:
1) In looking at the source code it appears that the cache size is just 32 for
this. From what I am
hearing from support every time I create a file for example it will have to
go through this code
path. If this is so, does it not seem that 32 is small? Also, it appears to
be hard coeded in the
source.
2) Also, is there a way to find out what where nfsauth is spending its time? I
am not familar with
this code stack enough to know what to probe. I would think dtrace would be
useful here but
I am not familar enough to know what to probe.
Any suggestions?
This message posted from opensolaris.org
