Merge lp://qastaging/~hopem/charms/trusty/ceph-radosgw/lp1517551 into lp://qastaging/~openstack-charmers-archive/charms/trusty/ceph-radosgw/next

charm_lint_check #14365 ceph-radosgw-next for hopem mp277882
    LINT OK: passed

Build: http://10.245.162.77:8080/job/charm_lint_check/14365/

charm_unit_test #13390 ceph-radosgw-next for hopem mp277882
    UNIT OK: passed

Build: http://10.245.162.77:8080/job/charm_unit_test/13390/

charm_amulet_test #8032 ceph-radosgw-next for hopem mp277882
    AMULET FAIL: amulet-test failed

AMULET Results (max last 2 lines):
Timeout occurred (2700s), printing juju status...environment: osci-sv08
ERROR:root:Make target returned non-zero.

Full amulet test output: http://paste.ubuntu.com/13501909/
Build: http://10.245.162.77:8080/job/charm_amulet_test/8032/

Rerunning amulet test...

charm_amulet_test #8052 ceph-radosgw-next for hopem mp277882
    AMULET FAIL: amulet-test failed

AMULET Results (max last 2 lines):
make: *** [functional_test] Error 1
ERROR:root:Make target returned non-zero.

Full amulet test output: http://paste.ubuntu.com/13581819/
Build: http://10.245.162.77:8080/job/charm_amulet_test/8052/

charm_amulet_test #8056 ceph-radosgw-next for hopem mp277882
    AMULET FAIL: amulet-test failed

AMULET Results (max last 2 lines):
make: *** [functional_test] Error 1
ERROR:root:Make target returned non-zero.

Full amulet test output: http://paste.ubuntu.com/13584216/
Build: http://10.245.162.77:8080/job/charm_amulet_test/8056/

charm_amulet_test #8057 ceph-radosgw-next for hopem mp277882
    AMULET FAIL: amulet-test failed

AMULET Results (max last 2 lines):
Timeout occurred (2700s), printing juju status...environment: osci-sv08
ERROR:root:Make target returned non-zero.

Full amulet test output: http://paste.ubuntu.com/13586064/
Build: http://10.245.162.77:8080/job/charm_amulet_test/8057/

charm_amulet_test #8064 ceph-radosgw-next for hopem mp277882
    AMULET FAIL: amulet-test failed

AMULET Results (max last 2 lines):
make: *** [functional_test] Error 1
ERROR:root:Make target returned non-zero.

Full amulet test output: http://paste.ubuntu.com/13600378/
Build: http://10.245.162.77:8080/job/charm_amulet_test/8064/

charm_lint_check #15769 ceph-radosgw-next for hopem mp277882
    LINT OK: passed

Build: http://10.245.162.77:8080/job/charm_lint_check/15769/

charm_unit_test #14717 ceph-radosgw-next for hopem mp277882
    UNIT OK: passed

Build: http://10.245.162.77:8080/job/charm_unit_test/14717/

charm_amulet_test #8383 ceph-radosgw-next for hopem mp277882
    AMULET OK: passed

Build: http://10.245.162.77:8080/job/charm_amulet_test/8383/

charm_unit_test #15485 ceph-radosgw-next for hopem mp277882
    UNIT OK: passed

Build: http://10.245.162.77:8080/job/charm_unit_test/15485/

charm_lint_check #16584 ceph-radosgw-next for hopem mp277882
    LINT OK: passed

Build: http://10.245.162.77:8080/job/charm_lint_check/16584/

charm_amulet_test #8500 ceph-radosgw-next for hopem mp277882
    AMULET OK: passed

Build: http://10.245.162.77:8080/job/charm_amulet_test/8500/

Thanks for the help, Edward! It looks like this merge proposal merges the two ports that the RadosGW uses currently, both the web server port and the civetweb port. Generally, when using civitweb, port 80 ends up load balanced between ALL RadosGW instances and port 70 is the port used to directly access the targeted machine. Maybe there should be two settings to direct that?

Chris

I think this update is OK - the way the port funnel works means that the configuration option is applied as the port which should be used to access the service, whether in hamode or node - Ed's change just makes this configurable - so if the normal 80 port is used then haproxy listens on 80, civetweb or apache on 70.

Likewise for 8080 - haproxy on 8080, civetweb on 8070.

As the amulet test is passing, we're not regressing function here, so +1 for me on this MP.

I'm only concerned that we seem to be merging port 70 and port 80 in this change. It doesn't kill me as accessing port 70 currently is only useful if trying to reduce interference in performance testing and shouldn't be done in a production setting.

> On 9 Jan 2016, at 15:32, Chris MacNaughton <email address hidden> wrote:
>
> I'm only concerned that we seem to be merging port 70 and port 80 in this change. It doesn't kill me as accessing port 70 currently is only useful if trying to reduce interference in performance testing and shouldn't be done in a production setting.

Maybe I missed something - I'll take another look

> --
> https://code.launchpad.net/~hopem/charms/trusty/ceph-radosgw/lp1517551/+merge/277882
> You are reviewing the proposed merge of lp:~hopem/charms/trusty/ceph-radosgw/lp1517551 into lp:~openstack-charmers/charms/trusty/ceph-radosgw/next.

These are the two different ports merging that I see

charm_lint_check #17076 ceph-radosgw-next for hopem mp277882
    LINT OK: passed

Build: http://10.245.162.77:8080/job/charm_lint_check/17076/

charm_unit_test #15951 ceph-radosgw-next for hopem mp277882
    UNIT OK: passed

Build: http://10.245.162.77:8080/job/charm_unit_test/15951/