0.9.5 release plans
Mark Reinhold
mr at sun.com
Wed Oct 17 23:48:16 CDT 2007
> Date: Wed, 17 Oct 2007 13:55:19 -0700
> From: danek.duvall at sun.com
> On Wed, Oct 17, 2007 at 04:21:58PM +0300, Giorgos Keramidas wrote:
>>> -abort: error: Connection refused
>>> +abort: error: Network is unreachable
>>
>> I think that's a difference between errno values returned by Solaris vs.
>> Linux vs. BSD. The same test `error' has been happening with Hg tests
>> on Solaris 10 for ages here.
>
> No, they're different errors. It's almost certainly that your IPv6 setup
> is incomplete. If you attempt to make a connection to "localhost", it'll
> first try the IPv4 address, 127.0.0.1, and get a connection refused error
> because nothing's listening on the port, so it falls back to the IPv6
> loopback address, ::1, and get the network unreachable, and errors out with
> that.
Aha, that is indeed what's happening.
> If you have IPv6 fully set up, then you should get "Connection refused" on
> both, and you won't see this diff. ...
>
> Alternatively, you should be able to disable ::1 by removing that line from
> /etc/hosts.
There is no ::1 line in /etc/hosts on this machine.
>> I don't know what is a good/portable way to fix this without trimming
>> useful information out of the error.
>
> Yeah, I don't know what the right answer here is. My guess is it's a
> busted system setup, and to require the end-user to get that right, even
> though such bustage is very common, and it's not clear what's actually
> going on.
I've checked this out on two different S10 installs (one fairly old, one
S10u3 (= 11/06)). I didn't enable ipv6 when installing these machines,
or fiddle with it later. If this is bustage then it's going to be very
common indeed, so it likely makes sense for the tests to code around it.
- Mark
More information about the Mercurial
mailing list