Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 24 Jan 2015 00:13:30 -0800
From:      perryh@pluto.rain.com (Perry Hutchison)
To:        m.seaman@infracaninophile.co.uk
Cc:        freebsd-questions@freebsd.org
Subject:   Re: how to get broken pkg build server fixed?
Message-ID:  <54c3542a.fxzPdaKPtvogn61R%perryh@pluto.rain.com>
In-Reply-To: <54C212D6.6030005@infracaninophile.co.uk>
References:  <54c205a1.nzckSmGXXPRQHMva%perryh@pluto.rain.com> <54C212D6.6030005@infracaninophile.co.uk>

next in thread | previous in thread | raw e-mail | index | archive | help
Matthew Seaman <m.seaman@infracaninophile.co.uk> wrote:
> On 01/23/15 08:26, Perry Hutchison wrote:
> > A week ago I got an email complaining of a package build failure
> > in a port I maintain, with the first reported error being:
> > 
> > !!! Jail is newer than host. (Jail: 1100052, Host: 1100050) !!!
> > !!! This is not supported. !!!
> > !!! Host kernel must be same or newer than jail. !!!
> > !!! Expect build failures. !!!
>
> This is something that portmgr and the build server admins will
> already be aware of.  While it looks alarming, in this case, it
> is in fact completely harmless.

So, in this case, it is a bug in Poudriere to scream about it like
that? 

How, in general, does one figure out whether this message matters in
a particular case?  (I figure it must sometimes be very important,
or the Poudriere folks would not have made it so prominent.)

> The 'Jail newer than host' thing would affect *every* package
> build on the cluster for FreeBSD HEAD ...

... unless the particular incompatibility introduced by a particular
instance of the misconfiguration turned out to affect only a few ports.

> Keep reading further down the email: usually the breakage
> you are responsible for looking at will occur near the end.

The later errors also seem to involve configuration problems --
it looks like a mismatch between the port's code and some .h file.

The port hasn't been touched in months -- at least, not to my
knowledge -- and it does not depend on other ports, so if the
Poudriere misconfiguration isn't the cause I guess that leaves
an incompatible change in base sometime within a week or so
before 1/11 (first Poudriere message regarding this, which
I didn't notice at the time).  That's not something I can
easily look into, because I don't run HEAD.

Now what?



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?54c3542a.fxzPdaKPtvogn61R%perryh>