openstack-qa-team team mailing list archive

[Dan Smith <danms@xxxxxxxxxx>]

JB> * Dependent patch series are tested in the correct order (regardless of
JB>   the order in which individual patches are approved, and patches are
JB>   only queued for testing once all the patches ahead of them are
JB>   approved).  This eliminated a huge amount of churn, as you can
JB>   imagine.

So, I'm not sure I'm seeing that exactly. I'm not entirely sure about
the mechanics behind gerrit and how it interacts with my git history,
but I see some patches later in the set going before the earlier
ones. This looks like it might be the order in which gerrit first saw
them in linear history (or something like that), but not the order in
which git knows them to be dependent. If I was smart and could write
everything in the right order the first time, then the orders would be
the same, but since I'm not... :)

I just verified this by looking at the zuul queue link you provided. I
see in the check queue things in the following order:

11411, 11500, 11823, 11754, 11440, 11351, 11353, 11482

but according to git, those should be in this order:

11411, 11351, 11352, 11353, 11410, 11440, 11481, 11482

Not a huge deal, of course, but getting that ordering right is key to
succeeding in the goal of not running dependent patches if their
ancestors fail.

Thanks!

-- 
Dan Smith
IBM Linux Technology Center