On Wed, 2020-06-17 at 08:34 -0500, Patrick Goetz wrote: > > mount_nfs_default_protocol = 4 > > to /etc/autofs.conf. However, using this setting in /etc/auto.home: > > * octopus.biosci.utexas.edu:/home/& > > any attempt to access a non-existent directory in /home (e.g. > /home/syslog) would result in the automounter hanging while various > attempts to mount were executed. Changing the auto.home line to > > * -tcp,vers=4.2 octopus.biosci.utexas.edu:/home/& > > resolved this issue. I can understand the tcp option, but what > doesn't > make sense to me is the necessity to also specify vers=4.2 > If I remove vers=4.2 from the options list, the automounter starts > hanging again when asked to mount a non-existent directory. If you specify vers or nfsvers and the its value starts with 4 (and mount_nfs_default_protocol is set to 4, and UDP protocol is not the requested mount protocol) then autofs will probe the server to work out whether it's available before attempting a mount. The point is, when the NFS version is explicitly given and those other conditions are met autofs will do targeted probing to work out if the server is responding in order to try to (partly) avoid the potential RPC stubbornness we see during the kernel mounting. If not the usual probing will be done for protocols and NFS versions to work out, if the server is responding, and what protocols and NFS versions are available. In fact it's not that smart, the probing is really only used to work out whether to attempt a mount rather than construct an options string, assuming that job is better left to mount.nfs(8). If there are multiple servers for the the mount entry (either via DNS or the map entry itself) then it does eliminate servers that aren't responding and order the list of servers to try by response time. Umm ... although I think there are also conditions under which it won't do the probe at all ... those bits are in more than one place, I'd need to look over that code if that's really going to be useful and needed. Ian
