Hi Paul,
the operating system is SLC 5 [ http://www.ScientificLinux.org ].
It's built on Red Hat sources à la CentOS (but with some customizations).
In this context you can consider it a Red Hat 5.
I didn't use the default versions of erlang and spidermonkey, instead I
used more recent versions that some folks here at work compiled (because
they packaged also the version 1.1.1 of couchdb):
# rpm -qa|egrep "^(erlang-R|js-)"
js-1.8.5-6
js-devel-1.8.5-6
erlang-R14B-03.3.el5
I tried to have a look at the failing test code
(test/etap/200-view-group-no-db-leaks.t) but I would have no idea how to
modify it in order to find out which test is failing (and I didn't find
any specific logs about it).
As reported in the original post, the user used is root, which I'm aware
being a bad idea but this was just to test that everything worked before
a real deploy.
Cheers, Elisiano.
On 04/01/2012 08:00 PM, Paul Davis wrote:
Awkward. This is a specific test that you can mostly ignore as we're
only using it to test the behavior of errors when views are opened.
Though I would be quite interested to know how that test fails for
you. Basically what happens is that we create a view, close it, remove
read permissions and then try and open it. I wonder if its perhaps not
closing quickly enough and you're getting a cached copy maybe?
Also, what OS and what user are you running the tests as?
On Fri, Mar 30, 2012 at 6:49 AM, Elisiano Petrini<[email protected]> wrote:
I forgot to include the detailed test log in the previous post.
It can be found here: http://paste.pocoo.org/show/573495/
Cheers, Elisiano
