)]}'
{"extraconfig/tasks/major_upgrade_pacemaker_migrations.sh":[{"author":{"_account_id":20172,"name":"Michele Baldessari","email":"michele@acksyn.org","username":"michele"},"change_message_id":"92ebd98d52ff8635288209255ae0a7c1ee27f3e7","unresolved":false,"context_lines":[{"line_number":25,"context_line":"    fi"},{"line_number":26,"context_line":"    # Only run this on the bootstrap node"},{"line_number":27,"context_line":"    if [ \"$(hiera -c /etc/puppet/hiera.yaml bootstrap_nodeid)\" \u003d \"$(facter hostname)\" ]; then"},{"line_number":28,"context_line":"        # Due to LP #1599798"},{"line_number":29,"context_line":"        pcs resource unmanage httpd-clone"},{"line_number":30,"context_line":"        CIB\u003d\"/root/liberty-cib.xml\""},{"line_number":31,"context_line":"        CIB_BACKUP\u003d\"/root/liberty-cib-orig.xml\""}],"source_content_type":"text/x-sh","patch_set":1,"id":"1aa78d24_9db5747f","line":28,"range":{"start_line":28,"start_character":8,"end_line":28,"end_character":28},"updated":"2016-07-07 11:41:49.000000000","message":"Best if we add some comments here like:\nWe unmanage the httpd resource to make sure that pacemaker won\u0027t race with the keystone deletion/stopping during the CIB transaction that will take place later","commit_id":"b031f6b2d11121a0a5c392caf544cbd719c1d103"},{"author":{"_account_id":20172,"name":"Michele Baldessari","email":"michele@acksyn.org","username":"michele"},"change_message_id":"92ebd98d52ff8635288209255ae0a7c1ee27f3e7","unresolved":false,"context_lines":[{"line_number":63,"context_line":"        $PCS resource delete openstack-keystone-clone"},{"line_number":64,"context_line":"        pcs cluster cib-push $CIB"},{"line_number":65,"context_line":""},{"line_number":66,"context_line":"        # manage httpd"},{"line_number":67,"context_line":"        pcs resource manage httpd-clone"},{"line_number":68,"context_line":"        # make sure httpd (which provides keystone now) are started after dummy"},{"line_number":69,"context_line":"        pcs constraint order start openstack-core-clone then httpd-clone"},{"line_number":70,"context_line":"        check_resource httpd started 1800"}],"source_content_type":"text/x-sh","patch_set":1,"id":"1aa78d24_1d586455","line":67,"range":{"start_line":66,"start_character":8,"end_line":67,"end_character":39},"updated":"2016-07-07 11:41:49.000000000","message":"I think we can move this after the last constraint change below. That way all the constraint changes are done before we manage the resource. Before the manage I think we should add a check to make sure the keystone resource is really gone before we manage the resource, I am not 100% convinced that after cib-push we are guaranteed that keystone is completely stopped and gone. Not sure we can use check_resource for that. Probably a simple pcs status |grep keystone should be sufficient (or a nicer variant thereof)","commit_id":"b031f6b2d11121a0a5c392caf544cbd719c1d103"},{"author":{"_account_id":8449,"name":"Marios Andreou","email":"marios.andreou@gmail.com","username":"marios"},"change_message_id":"8ff702f9c6f1428282257f9b3a6a7109ab7af9fe","unresolved":false,"context_lines":[{"line_number":66,"context_line":""},{"line_number":67,"context_line":"        # Let\u0027s be 100% sure that the keystone resource is stopped and gone before"},{"line_number":68,"context_line":"        # we remanage the httpd resource later below. We cannot reuse check_resource"},{"line_number":69,"context_line":"        # as the resource might not exist already in which case the function would fail"},{"line_number":70,"context_line":"        tstart\u003d$(date +%s)"},{"line_number":71,"context_line":"        while pcs status | grep -q keystone-clone; do"},{"line_number":72,"context_line":"            sleep 5"}],"source_content_type":"text/x-sh","patch_set":2,"id":"1aa78d24_17100aa8","line":69,"updated":"2016-07-08 10:12:15.000000000","message":"\"might not exist already\" you mean may have already been deleted right? in any case lets be explicit here and do what you have below - sounds like we could fix it up on the check_resource side (exist grafefully if it doesn\u0027t exist) but is fine for now if this works and then we can tidy/iterate.","commit_id":"84019ac86c6329132d0f728acb427684ea838e7d"},{"author":{"_account_id":8042,"name":"Jiří Stránský","email":"jistr@redhat.com","username":"jistr"},"change_message_id":"abc895ad520dd9646414cc88ce1dfac501176192","unresolved":false,"context_lines":[{"line_number":66,"context_line":""},{"line_number":67,"context_line":"        # Let\u0027s be 100% sure that the keystone resource is stopped and gone before"},{"line_number":68,"context_line":"        # we remanage the httpd resource later below. We cannot reuse check_resource"},{"line_number":69,"context_line":"        # as the resource might not exist already in which case the function would fail"},{"line_number":70,"context_line":"        tstart\u003d$(date +%s)"},{"line_number":71,"context_line":"        while pcs status | grep -q keystone-clone; do"},{"line_number":72,"context_line":"            sleep 5"}],"source_content_type":"text/x-sh","patch_set":2,"id":"1aa78d24_f60ed230","line":69,"in_reply_to":"1aa78d24_17100aa8","updated":"2016-07-11 09:25:41.000000000","message":"+1 on keeping this implementation.\n\nIf such a thing comes up in the future we could change check_resource to support more than just stopped/started, e.g. \u0027deleted\u0027. I\u0027d be -1 on changing `check_resource stopped` to exit without error if resource isn\u0027t defined -- we could miss some errors/typos that way.","commit_id":"84019ac86c6329132d0f728acb427684ea838e7d"}]}