Re: [Gluster-devel] FreeBSD and NetBSD builds failing on master

[Jeff Darcy]

On November 2, 2015 at 12:09:45 PM, Emmanuel Dreyfus (m...@netbsd.org) wrote:
> On Mon, Nov 02, 2015 at 11:50:25AM -0500, Jeff Darcy wrote:
> > I should have caught this problem (mea culpa) but ?break the world? is a
> > bit hyperbolic. ?The world? is broken much of the time, because very
> > few of the developers on this project know how to fix *BSD portability
> > problems. How would they, when nobody ever tries to educate them in
> > review comments or elsewhere?
>  
> Well, I never refused ot help on that front. I just cannot keep up
> with the change stream, so people just have to ask.

None of us can keep up with the change stream.  :(  What’s the best way
to ask, in your case?  Add you as a reviewer within Gerrit?  Separate
email?  Anything else that would maximize efficiency when people ask for
help?

Also, you’re still only one person whose responsibilities include much
more than Gluster.  Several others have picked up bits and pieces of
knowledge about how to make sure things work on *BSD.  How do we make
sure they are also aware of “BSD blockers” in areas they might not be
watching?  Would a separate Gerrit flag for BSD verification help?  Or a
gluster-bsd mailing list?
___
Gluster-devel mailing list
Gluster-devel@gluster.org
http://www.gluster.org/mailman/listinfo/gluster-devel

Re: [Gluster-devel] FreeBSD and NetBSD builds failing on master

[Vijay Bellur]

On Monday 02 November 2015 03:41 PM, Niels de Vos wrote:

On Mon, Nov 02, 2015 at 01:19:48PM +0530, Raghavendra Talur wrote:

Hi All,

Recently merged patch "debug/io-stats: Add FOP sampling feature" [1] has a
hard coded path for log directory and I think *BSD machines have a
different path for /var/log.

Can we have the patch reverted till this is fixed?


[1] http://review.gluster.org/#/c/12210/


We really need to be careful not to merge patches that have not passed
the NetBSD/FreeBSD smoke tests. The comments clearly state FAILURE, and
the NetBSD regression test was not passed either yet.



mea culpa here. It was an oversight from me and I  will ensure that the 
*BSD votes are not overridden without adequate due diligence before a merge.




http://review.gluster.org/12484 is the fix, it has passed all smoke
tests. If there is a real urgency, we can merge that before the
regression tests have passed.



I have merged this patch ignoring the spurious regression test failure.

Thanks,
Vijay


___
Gluster-devel mailing list
Gluster-devel@gluster.org
http://www.gluster.org/mailman/listinfo/gluster-devel

Re: [Gluster-devel] FreeBSD and NetBSD builds failing on master

[Emmanuel Dreyfus]

Jeff Darcy  wrote:

> None of us can keep up with the change stream.  :(  What's the best way
> to ask, in your case?  Add you as a reviewer within Gerrit?  Separate
> email?  Anything else that would maximize efficiency when people ask for
> help?

Well an e-mail is fine. Being on the review is a bit unhelpful since
that does not explain the problem encountered.
 
> Also, you're still only one person whose responsibilities include much
> more than Gluster.  Several others have picked up bits and pieces of
> knowledge about how to make sure things work on *BSD.  How do we make
> sure they are also aware of "BSD blockers" in areas they might not be
> watching?  Would a separate Gerrit flag for BSD verification help?  Or a
> gluster-bsd mailing list?

Well, the regression tests are supposed to catch problems, and I can
help fixing them if people ask for help.

-- 
Emmanuel Dreyfus
http://hcpnet.free.fr/pubz
m...@netbsd.org
___
Gluster-devel mailing list
Gluster-devel@gluster.org
http://www.gluster.org/mailman/listinfo/gluster-devel

Re: [Gluster-devel] FreeBSD and NetBSD builds failing on master

[Jeff Darcy]

> Generally speaking, I am in favor of backing otu changes that
> break the world.

I should have caught this problem (mea culpa) but “break the world” is a
bit hyperbolic.  “The world” is broken much of the time, because very
few of the developers on this project know how to fix *BSD portability
problems.  How would they, when nobody ever tries to educate them in
review comments or elsewhere?  If we just say that *BSD failures block
all merges, but nobody tries to speed the process of fixing *BSD
failures, then that slows down overall development too much.  As I
understand it, that’s why those results currently don’t count toward
verification in Gerrit.

Some of us have talked about instituting a delay or window in which
patches that only fail on *BSD can be fixed up for that environment,
without obstructing other development.  If anyone has ideas on how to
make that work, or has a different idea that doesn’t make *BSD the
bottleneck for all development on all platforms, now would be a great
time to share those ideas.
___
Gluster-devel mailing list
Gluster-devel@gluster.org
http://www.gluster.org/mailman/listinfo/gluster-devel

Re: [Gluster-devel] FreeBSD and NetBSD builds failing on master

[Emmanuel Dreyfus]

On Mon, Nov 02, 2015 at 01:19:48PM +0530, Raghavendra Talur wrote:
> Recently merged patch "debug/io-stats: Add FOP sampling feature" [1] has a
> hard coded path for log directory and I think *BSD machines have a
> different path for /var/log.

On NetBSD /var/log is the location for logs, but one have to 
create the subdirectories.

> Can we have the patch reverted till this is fixed?

Generally speaking, I am in favor of backing otu changes that
break the world.

-- 
Emmanuel Dreyfus
m...@netbsd.org
___
Gluster-devel mailing list
Gluster-devel@gluster.org
http://www.gluster.org/mailman/listinfo/gluster-devel

[Gluster-devel] FreeBSD and NetBSD builds failing on master

[Raghavendra Talur]

Hi All,

Recently merged patch "debug/io-stats: Add FOP sampling feature" [1] has a
hard coded path for log directory and I think *BSD machines have a
different path for /var/log.

Can we have the patch reverted till this is fixed?


[1] http://review.gluster.org/#/c/12210/

Thanks,
Raghavendra Talur
___
Gluster-devel mailing list
Gluster-devel@gluster.org
http://www.gluster.org/mailman/listinfo/gluster-devel