08:57:34 <szilard> I just learned about the KVM zone brand. Do the brand of a HVM zone have any significant performance impact?
09:35:45 <m1ari> for a new HVM zone you may do better to look at bhyve rather than kvm.
18:12:04 <patrikr> whats the proper format to put an ipv6 subnet as a sharenfs option? i enclose the address in brackets but it still seems to split by colon
18:30:54 <jclulow> patrikr: Are you trying to use a prefix length, like /48
18:31:16 <patrikr> yes
18:34:08 <jclulow> Are you doing like [fe80::1]/48 or [fe80::1/48]
18:34:31 <patrikr> the former
18:34:31 <jclulow> Also are you putting the @ on the front
18:54:29 <jclulow> patrikr: So, looking at in_access_list() in usr/src/cmd/fs.d/nfs/mountd/mountd.c it seems like the syntax is essentially...   sharenfs=@[fe80::1]/48
18:54:33 <jclulow> patrikr: So, looking at in_access_list() in usr/src/cmd/fs.d/nfs/mountd/mountd.c it seems like the syntax is essentially...   sharenfs=rw=@[fe80::1]/48
18:54:36 <jclulow> sigh
18:54:47 <jclulow> the second one haha
18:55:36 <jclulow> The @ is what tells it that it could be an IP address
18:56:03 <patrikr> right, i suppose showmount just shows it incorrectly then?
18:56:27 <jclulow> It's possible!  What do you see from showmount?
18:57:26 <patrikr> "@[fde8,cafe,aaaa,]/48," on both the server and my linux desktop
18:57:50 <jclulow> What about "sharemgr show -p"
18:58:12 <patrikr> shows up properly there
18:59:15 <jclulow> Ok, that's good.  showmount(8) uses RPC requests to mountd(8), and it sounds like maybe there is a bug in what mountd is rendering for that option
19:00:02 <jclulow> If you could file that bug against illumos that'd be great
19:00:27 <jclulow> (it seems like mountd because you see the same thing on your Linux machine)
19:01:43 <patrikr> i could, just on bugs.illumos.org?
19:02:05 <jclulow> Yeah
19:02:20 <jclulow> thanks!
21:18:33 <patrikr> made the bug report now