On Sun, 2012-12-30 at 19:04 -0800, H. Peter Anvin wrote: > On 12/30/2012 06:59 PM, H. Peter Anvin wrote: > > On 12/19/2012 06:31 PM, H. Peter Anvin wrote: > >> On 12/19/2012 05:52 PM, Ian Kent wrote: > >>>> > >>>> I'll check what I have in f17 but in the meantime, if you want to, you > >>>> could try rebuilding the rpm from f18 or Rawhide, they're the same at > >>>> this point. > >>> > >>> I see the getifaddrs(3) isn't in f17 but the simple mount probing is so > >>> that could be the problem. Building he later package and checking if > >>> that works should verify if that is the problem. Though I would still > >>> need to check if there are other recent changes that make a difference. > >>> > >> > >> I rebuilt autofs-5.0.7-9 from the fc18 repo for fc17; the problem > >> remains and is identical. > >> > > > > OK, I think I figured out the problem -- and it *is* a problem, just not > > the one I expected. > > > > The fileserver had a bad AAAA record; the A record was correct. autofs, > > unlike mount, didn't fall back to IPv4 (this is part of why I really > > wanted this logic to be outside of autofs back in the autofs3 days, but > > that's ancient history.) So automounting by IPv4 address worked. That > > automounting by IPv6 address didn't work is more confusing, I think -- > > and it doesn't, with or without brackets. Yeah, that could be changed to exec showmount but the RPC code used for this is the same that's used to probe availability (essentially a NULL proc ping). The availability probe is needed all the time now now because, with mount.nfs letting the kernel do most of the work, there can be significant delays if a server isn't available. Unfortunately the RPC code needs to be fairly low level as well to control timeouts to give reasonable interactive response. So this does need to be sorted out. I was thinking that it might be the error handling in the RPC code bailing out to early and not trying all addresses. Is the mountd server service also available and working over IPv6? For simple indirect (and direct) mounts you can prevent the probe being done by setting MOUNT_WAIT to a sensible value, say between 15 and 30 seconds, depending on your preference. Checking that will eliminate much of the parsing code as part of the problem, although it does sound like this is the RPC code anyway. > > > > Hm, no, even with the correct AAAA record it doesn't work via IPv6. I tried to duplicate it but don't think I got the environment right. I'll need have another go at setting up an IPv6 environment, including DNS, to check it. Ian -- To unsubscribe from this list: send the line "unsubscribe autofs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
