)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":12279,"name":"Janie Richling","email":"jrichli@us.ibm.com","username":"jrichli"},"change_message_id":"82d6a5d895dc6e27b34ce82486c9083fd0fa20a9","unresolved":false,"context_lines":[{"line_number":20,"context_line":"failed, since the condition that would have skipped the test was never"},{"line_number":21,"context_line":"met."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"Change-Id: I836294256d47b71ea23fa3b758b72f4b46b93db3"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"9a8ffd7b_714db471","line":23,"updated":"2015-11-24 19:58:03.000000000","message":"Please add the \"Closes-Bug\" for https://bugs.launchpad.net/swift/+bug/1517974 so that this will be tracked together.\n\nThis is explained in https://wiki.openstack.org/wiki/GitCommitMessages#Including_external_references","commit_id":"cff0451ae8ceb6e09ed1b39bc471e034e9f7a16a"}],"test/functional/tests.py":[{"author":{"_account_id":15343,"name":"Tim Burke","email":"tburke@nvidia.com","username":"tburke"},"change_message_id":"855c80fdf09098e52ece09bd78f135f8289bca13","unresolved":false,"context_lines":[{"line_number":2684,"context_line":"        cls.storage_url, cls.storage_token \u003d cls.conn.authenticate()"},{"line_number":2685,"context_line":""},{"line_number":2686,"context_line":"        if cls.versioning_enabled is None:"},{"line_number":2687,"context_line":"            cls.versioning_enabled \u003d \u0027versioned_writes\u0027 in cluster_info"},{"line_number":2688,"context_line":"            if not cls.versioning_enabled:"},{"line_number":2689,"context_line":"                return"},{"line_number":2690,"context_line":""}],"source_content_type":"text/x-python","patch_set":1,"id":"9a8ffd7b_ab62cdd6","line":2687,"updated":"2015-11-24 19:24:32.000000000","message":"Versioning can still be enabled despite versioned_writes not being in cluster_info (or even in the explicit pipeline at all). Prior to versioned_writes being middleware, the container-server settings (!) dictated whether or not X-Versions-Location would persist.\n\nI think the better fix for this is to catch the 412 down around line 2706 and use that as the sign that versioning tests should be skipped.","commit_id":"cff0451ae8ceb6e09ed1b39bc471e034e9f7a16a"},{"author":{"_account_id":12279,"name":"Janie Richling","email":"jrichli@us.ibm.com","username":"jrichli"},"change_message_id":"ad7793f7b28a2a8de3f9ee4813c935af07340176","unresolved":false,"context_lines":[{"line_number":2684,"context_line":"        cls.storage_url, cls.storage_token \u003d cls.conn.authenticate()"},{"line_number":2685,"context_line":""},{"line_number":2686,"context_line":"        if cls.versioning_enabled is None:"},{"line_number":2687,"context_line":"            cls.versioning_enabled \u003d \u0027versioned_writes\u0027 in cluster_info"},{"line_number":2688,"context_line":"            if not cls.versioning_enabled:"},{"line_number":2689,"context_line":"                return"},{"line_number":2690,"context_line":""}],"source_content_type":"text/x-python","patch_set":2,"id":"9a8ffd7b_00a0869e","line":2687,"updated":"2015-11-25 03:43:59.000000000","message":"Tim\u0027s comment still needs to be addressed, correct?  If we assume that a 412 implies that we should skip, could that mask new issues?  I am wondering if we could check the container-server settings also to detect an old setup.  Is an old setup relevant for this change to functional tests?  If somebody has a setup without versioned_writes as middleware, would they get the new functests and run them against the cluster?","commit_id":"c91e1d4b56731719437830a6ef588cac7dcc6666"},{"author":{"_account_id":15343,"name":"Tim Burke","email":"tburke@nvidia.com","username":"tburke"},"change_message_id":"d1534c953bca235140e523f6bb442cbbf1cf141e","unresolved":false,"context_lines":[{"line_number":2684,"context_line":"        cls.storage_url, cls.storage_token \u003d cls.conn.authenticate()"},{"line_number":2685,"context_line":""},{"line_number":2686,"context_line":"        if cls.versioning_enabled is None:"},{"line_number":2687,"context_line":"            cls.versioning_enabled \u003d \u0027versioned_writes\u0027 in cluster_info"},{"line_number":2688,"context_line":"            if not cls.versioning_enabled:"},{"line_number":2689,"context_line":"                return"},{"line_number":2690,"context_line":""}],"source_content_type":"text/x-python","patch_set":2,"id":"9a8ffd7b_37970a92","line":2687,"in_reply_to":"9a8ffd7b_00a0869e","updated":"2015-11-25 22:51:21.000000000","message":"Checking container-server settings may be tricky. My understanding was that we\u0027re only really given an auth endpoint and some creds, such as in https://github.com/openstack/swift/blob/master/test/sample.conf -- from that we can eventaully make /info requests, but I don\u0027t think we can really probe container-server settings.\n\nJust to clarify: \"old setup\" in this case means \"old config\", not \"old code\". You can have a current proxy-server without versioned_writes explicitly in the pipeline [1]; on startup, the proxy manipulates its own pipeline to ensure it\u0027s in there, similar to what\u0027s done with gatekeeper and dlo.\n\nMaybe that\u0027s beside the point, though. I could certainly be convinced that we don\u0027t need the functional tests to cover out-of-date configs.\n\n[1] Like, say, how vagrant-swift-all-in-one is currently configured: https://github.com/swiftstack/vagrant-swift-all-in-one/blob/master/cookbooks/swift/files/default/etc/swift/proxy-server/proxy-server.conf.d/20_settings.conf","commit_id":"c91e1d4b56731719437830a6ef588cac7dcc6666"}]}