-
jbkit's been a long time since i've messed with samba (and i've not actually tried it), but couldn't the smbautohome feature do that? -- setup a map like '* /home/&' so \\server\FOO shares out /home/foo if it exists (or could make it more intricate if desired)?
-
jbkerr not tried the smbautohome bits
-
jbk(mostly just wondering if there's things that should be added to the smb server)
-
jbki still need to figure out what to do with my mdns changes
-
jbkspeaking of that
-
jbki was thinking of maybe a slightly more general approach for that (which would avoid adding directly dependencies to mdns)
-
jbksomething similar in spirit to sc.ipfd where manifests could specify mDNS entries to broadcast if a service is enabled
-
jbkand maybe a script that could handle dynamic bits (e.g. the _adisk entry)
-
jbki guess linux clients can advertise _nfs services (though it's not clear if anything consumes them)
-
jclulowI think that seems reasonable; definitely IPD territory though
-
jclulowjbk: huh, TIL smbautohome(5)
-
jclulowmaybe it actually work haha
-
jclulow*would work
-
jclulowI have literally never seen this manual page
-
jbki mean i've not used it myself, so it could totally suck, but just curious for reasons people aren't using it -- if it's stuff that can be added or fixed, good to know what those are
-
jbkerr aren't using the in-kernel smb
-
jbki'd much rather it be 'oh i'm just used to samba' or whatever vs. 'smbd doesn't do what i need'
-
tsoomeI have been using smbautohome for years:D
-
sommerfeldtsoome: what sort of configuration do you have in your smbautohome?
-
tsoome"* /export/home/&"
-
tsoomevery trivial:)
-
sommerfeldTrivial is good.
-
jclulowDoes it have to point at /export/home or can it point at an existing autofs mount like /home
-
sommerfeldFYI I filed illumos.org/issues/16438
-
fenix→ BUG 16438: NFS4 xdr_share_access() fails to handle XDR_ENCODE path (New)
-
tsoomesommerfeld nice work:)
-
sommerfeldtsoome: Udo did the hard part of this.
-
sommerfeldwhich told me where to look (the code dealing with argument marshalling around this "open_downgrade" thing) and what to look for..
-
jclulowthanks for filing sommerfeld
-
gitomat[illumos-gate] 16225 cputrack should not be isaexec'd -- Richard Lowe <richlowe⊙rn>
-
gitomat[illumos-gate] 16434 kernel modules should use consistent relative source paths -- Richard Lowe <richlowe⊙rn>