On Fri, Mar 12, 1999 at 09:44:15PM -0800, Bill Studenmund wrote:
> Say I have a file for which a VOP_READ will take five minutes to complete
> (I have to restore some of its contents from tape).
> 
> Say this file lives in an nfs-exported fs.
> 
> How badly would this break an nfs server? The first nfsd to read the file
> would sleep for five minutes. What would happen then? Would the client
> retry, and possibly send another process to sleep (possibly getting all of
> them)?

Yes, this might happen.

> Would having the nfs server execute all its reads & writes with IO_NDELAY
> (and teaching my stuff to return EWOULDBLOCK) make sense?

I guess.. though you seem to having to go through an awful lot of trouble
modifying other parts of the kernel because of the sometimes non-standard
behavior of this.

- Frank