)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":7233,"name":"Matthew Oliver","email":"matt@oliver.net.au","username":"mattoliverau"},"change_message_id":"4a5736b86e7c814843ed3a79168b526daeab3087","unresolved":false,"context_lines":[{"line_number":6,"context_line":""},{"line_number":7,"context_line":"Remove redundant container updating after rsync"},{"line_number":8,"context_line":""},{"line_number":9,"context_line":"In today\u0027s swift, information abount reporting in container db"},{"line_number":10,"context_line":"get reset after rsync, and container-updater tries to update"},{"line_number":11,"context_line":"its account again from new position."},{"line_number":12,"context_line":"However, when a container db is rsynced after its account is"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":5,"id":"9a41bdd9_916c6836","line":9,"updated":"2015-07-15 06:52:35.000000000","message":"s/abount/about/","commit_id":"01e6364373c3c8e934c2717f37069bed291b3d00"}],"swift/common/db_replicator.py":[{"author":{"_account_id":1179,"name":"Clay Gerrard","email":"clay.gerrard@gmail.com","username":"clay-gerrard"},"change_message_id":"1151657e14d7bf2f71485cdf131ed75216c57d04","unresolved":false,"context_lines":[{"line_number":713,"context_line":"            return HTTPNotFound()"},{"line_number":714,"context_line":"        if not os.path.exists(old_filename):"},{"line_number":715,"context_line":"            return HTTPNotFound()"},{"line_number":716,"context_line":"        broker \u003d self.broker_class(old_filename)"},{"line_number":717,"context_line":"        broker.newid(args[0])"},{"line_number":718,"context_line":"        renamer(old_filename, db_file)"},{"line_number":719,"context_line":"        return HTTPNoContent()"}],"source_content_type":"text/x-python","patch_set":4,"id":"ba3cc151_432bf130","line":716,"updated":"2015-06-29 19:53:00.000000000","message":"doesn\u0027t the bug reference *this* one?  I really don\u0027t know why this would newid honestly...","commit_id":"cab04fea3cf73c83dcf1b06c6e3b54a2c58cd7d0"},{"author":{"_account_id":1179,"name":"Clay Gerrard","email":"clay.gerrard@gmail.com","username":"clay-gerrard"},"change_message_id":"1151657e14d7bf2f71485cdf131ed75216c57d04","unresolved":false,"context_lines":[{"line_number":731,"context_line":"            point \u003d objects[-1][\u0027ROWID\u0027]"},{"line_number":732,"context_line":"            objects \u003d existing_broker.get_items_since(point, 1000)"},{"line_number":733,"context_line":"            sleep()"},{"line_number":734,"context_line":"        new_broker.newid(args[0], False)"},{"line_number":735,"context_line":"        renamer(old_filename, db_file)"},{"line_number":736,"context_line":"        return HTTPNoContent()"},{"line_number":737,"context_line":""}],"source_content_type":"text/x-python","patch_set":4,"id":"ba3cc151_c3b6213d","line":734,"updated":"2015-06-29 19:53:00.000000000","message":"so the idea is that we can \"trust\" the other broker that shipped it\u0027s database with all these rows have already updated them to the account?\n\nBut it\u0027s not clear to me that\u0027s always safe?\n\nCould we check the reported stats in the new_broker - it\u0027s possible that even after eating all these new rows there\u0027s nothing we have that would invalidate what\u0027s already been reported by the remote?  (e.g. we some lazy old replica woke up from a offline node after ages of being out of the loop - so some remote ships us a db with 100K new rows that are more relevant than anything and everything we have - which it already had anyway - and now that we\u0027re on the same page our report stats are going to come up with exactly what\u0027s already been reported ages ago)","commit_id":"cab04fea3cf73c83dcf1b06c6e3b54a2c58cd7d0"},{"author":{"_account_id":1179,"name":"Clay Gerrard","email":"clay.gerrard@gmail.com","username":"clay-gerrard"},"change_message_id":"5146b7fad7fa047713bc0b67fa5043d65f6dea39","unresolved":false,"context_lines":[{"line_number":758,"context_line":"        if not os.path.exists(old_filename):"},{"line_number":759,"context_line":"            return HTTPNotFound()"},{"line_number":760,"context_line":"        broker \u003d self.broker_class(old_filename)"},{"line_number":761,"context_line":"        broker.newid(args[0], False)"},{"line_number":762,"context_line":"        renamer(old_filename, db_file)"},{"line_number":763,"context_line":"        return HTTPNoContent()"},{"line_number":764,"context_line":""}],"source_content_type":"text/x-python","patch_set":5,"id":"3a50d1a3_2e00f300","line":761,"updated":"2015-07-21 01:04:52.000000000","message":"rather than all the plumbing can we just *not* call newid when we rsync the whole db?","commit_id":"01e6364373c3c8e934c2717f37069bed291b3d00"}]}