Review fstab options for nfs clients.

[ken@…]

On current systemd, automatic mounting of nfs shares needs a workaround option in the fstab (or else root has to use mount -a), because current systemd tries to mount them before the network is up.

I was going to add this to the book, but then I realised that the existing examples are for nfs v2 and I suspect they are out of date. I only use nfs v3 and my options are somewhat different. In particular, I have had to specify vers=3 for some time.

Initial post at ​http://lists.linuxfromscratch.org/pipermail/blfs-dev/2014-April/027288.html, this ticket is to remind me to look at all the options, and to document them. Probably also need to look at current kernel options, both for the client and perhaps also for the server.

I can hack v3, and given time I can probably sort out v2 (I suspect that a version specification is also needed there too), but any advice from people using v4 would be welcome.

[ken@…]

In fact, much of what I have been specifying is not required - with recent linux systems, most things are negotiated. My server offers nfs v3 and uses tcp, removing those options from my fstab makes an unnoticeable change in the time taken to mount filesystems. My boxes negotiate:

vers=3,rsize=1048576,wsize=1048576,namelen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys

I'll update the book to simplify what is specified, and to add some explanations.

[ken@…]

r13180