Bugzilla – Bug 151618
dns resolver doesnt gracefully handle mdns resolver failures by falling back to dns
Last modified: 2011-04-28 08:42:37 UTC
from a SuSE10 (and earliser) attempting to resolve pingable (by IP) hosts with an fqdn ending in .local that are *not* running a properly configured msdnsd will always fail. the only workaround is to turn off mdns completely by adding mdns off to /etc/host.conf as documented in the following novell cool solution http://www.novell.com/coolsolutions/tip/15248.html this is not a desirable solution. the resolver should automagically fallback to dns if the mdns resolv fails. Why is this important? Because MSFT Best Practices suggest using .local (ie bob.noodle.local, server.foo.local) for a network that is not going to be exposed to the internet. We are seeing many potential customers for our product (Centeris Likewise) are running their networks in a mannner consistent with this convention. As a result, none of the SuSE boxes that are installed into these networks is capable of being a functioning member of these networks. In the latest update to our product (Centeris LikeWise 1.03) we will be setting mdns off in /etc/resolv.conf as part of our install process. this is a sad thing, we would prefer to leave the mdns functionality available for our users. it's possible that there is a further configuration option that we are not aware of that would make .local domains resolvable first by mdns and then if that fails then dns, but we dont know what it is. if you know of a solution to this problem that doesnt cripple mdns, please enter it into this bug, if not, please fix the resolver to do a better job in this situation. tnx! john utz, developer guy at centeris.com www.centeris.com
*** Bug 151615 has been marked as a duplicate of this bug. ***
Thorsten: Can you provide a comment here, please?
MDNS takes quite some time, so even if we add fallback code users will complain that resolving .local domain takes ages. I prefer an error here, as the user will then now that something fishy is going on.
Is this bug fixed? - I just vpn'ed into our network at work and remote machines ending in .local responded just fine ( it might be that our msdnsd is configured correctly now, but this used to fail for me) Also bug 130561 seems to be closely related to this one.
Closing with WONTFIX. (maybe FIXED)