)]}'
{"playbooks/roles/galera_server/tasks/galera_upgrade_post.yml":[{"author":{"_account_id":7219,"name":"git-harry","email":"git-harry@live.co.uk","username":"git-harry"},"change_message_id":"cbd179709e00d213d3a632dd6ede2f869f36d5b6","unresolved":false,"context_lines":[{"line_number":12,"context_line":"# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied."},{"line_number":13,"context_line":"# See the License for the specific language governing permissions and"},{"line_number":14,"context_line":"# limitations under the License."},{"line_number":15,"context_line":"- name: Start mariadb"},{"line_number":16,"context_line":"  service:"},{"line_number":17,"context_line":"    name: mysql"},{"line_number":18,"context_line":"    state: started"}],"source_content_type":"text/x-yaml","patch_set":1,"id":"dada55a8_cc75d2cb","line":15,"updated":"2016-07-26 09:18:46.000000000","message":"I can\u0027t see any comment as to why this change is being made. It seems separate from the rest of the patch which is concerned with restarting containers in a safe manner.","commit_id":"5ad9631366b13d782f1c784e0d24fa103c206df0"},{"author":{"_account_id":425,"name":"Darren Birkett","email":"darren.birkett@gmail.com","username":"darren-birkett"},"change_message_id":"3b086b6fcab3b143e490b8cef7e97306629a9d68","unresolved":false,"context_lines":[{"line_number":12,"context_line":"# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied."},{"line_number":13,"context_line":"# See the License for the specific language governing permissions and"},{"line_number":14,"context_line":"# limitations under the License."},{"line_number":15,"context_line":"- name: Start mariadb"},{"line_number":16,"context_line":"  service:"},{"line_number":17,"context_line":"    name: mysql"},{"line_number":18,"context_line":"    state: started"}],"source_content_type":"text/x-yaml","patch_set":1,"id":"dada55a8_07a86576","line":15,"in_reply_to":"dada55a8_cc75d2cb","updated":"2016-07-26 09:33:59.000000000","message":"This odd way of starting mysql was introduced because after an upgrade, \u0027sometimes mariadb crashed\u0027. This crashing was being caused by the uncontrolled container restarts. Now that is fixed, we can move back to a standard way of starting mysql.","commit_id":"5ad9631366b13d782f1c784e0d24fa103c206df0"}],"releasenotes/notes/mariadb-rolling-upgrades-323510425c3c7751.yaml":[{"author":{"_account_id":6816,"name":"Jesse Pretorius","email":"jesse@odyssey4.me","username":"jesse-pretorius"},"change_message_id":"e87ed62040e1c28b5662016905880853249ba844","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"upgrade:"},{"line_number":3,"context_line":"  - During a kilo to liberty upgrade, container and service restarts for the mariadb/galera cluster were being triggered multiple times and causing the cluster to become unstable and often unrecoverable. This situation has been improved immensely, and we now have tight control such that restarts of the galera containers only need to happen once, and are done so in a controlled, predictable and repeatable way."}],"source_content_type":"text/x-yaml","patch_set":3,"id":"dada55a8_314ccbc7","line":3,"updated":"2016-07-26 08:58:12.000000000","message":"Can you break the line at \u003c80 chars please?","commit_id":"a78ad66c3f48786baea1ec007a20128837f7bdd6"}],"scripts/run-upgrade.sh":[{"author":{"_account_id":7219,"name":"git-harry","email":"git-harry@live.co.uk","username":"git-harry"},"change_message_id":"cbd179709e00d213d3a632dd6ede2f869f36d5b6","unresolved":false,"context_lines":[{"line_number":156,"context_line":"        # explicitly perform mariadb upgrade"},{"line_number":157,"context_line":"        RUN_TASKS+\u003d(\"galera-install.yml -e \u0027galera_upgrade\u003dtrue\u0027\")"},{"line_number":158,"context_line":"        # explicitly perform controlled galera cluster restart"},{"line_number":159,"context_line":"        RUN_TASKS+\u003d(\"${UPGRADE_PLAYBOOKS}/galera-cluster-rolling-restart.yml\")"},{"line_number":160,"context_line":"        RUN_TASKS+\u003d(\"${UPGRADE_PLAYBOOKS}/repo-server-pip-conf-removal.yml\")"},{"line_number":161,"context_line":"        # individually run each of the remaining plays from setup-infrastructure"},{"line_number":162,"context_line":"        RUN_TASKS+\u003d(\"haproxy-install.yml\")"}],"source_content_type":"text/x-sh","patch_set":1,"id":"dada55a8_0ced2ab7","line":159,"updated":"2016-07-26 09:18:46.000000000","message":"Why not run this after the containers are modified. The galera playbook manages restarts itself, galera-cluster-rolling-restart seems to be for use in restarting the containers in a safe way. Since this restart is independent of the restart caused by the galera upgrade, having this happen a soon as the containers are changed keeps the change (modifying container config) as close as possible to the activation of that change (restart of the containers).","commit_id":"5ad9631366b13d782f1c784e0d24fa103c206df0"},{"author":{"_account_id":425,"name":"Darren Birkett","email":"darren.birkett@gmail.com","username":"darren-birkett"},"change_message_id":"3b086b6fcab3b143e490b8cef7e97306629a9d68","unresolved":false,"context_lines":[{"line_number":156,"context_line":"        # explicitly perform mariadb upgrade"},{"line_number":157,"context_line":"        RUN_TASKS+\u003d(\"galera-install.yml -e \u0027galera_upgrade\u003dtrue\u0027\")"},{"line_number":158,"context_line":"        # explicitly perform controlled galera cluster restart"},{"line_number":159,"context_line":"        RUN_TASKS+\u003d(\"${UPGRADE_PLAYBOOKS}/galera-cluster-rolling-restart.yml\")"},{"line_number":160,"context_line":"        RUN_TASKS+\u003d(\"${UPGRADE_PLAYBOOKS}/repo-server-pip-conf-removal.yml\")"},{"line_number":161,"context_line":"        # individually run each of the remaining plays from setup-infrastructure"},{"line_number":162,"context_line":"        RUN_TASKS+\u003d(\"haproxy-install.yml\")"}],"source_content_type":"text/x-sh","patch_set":1,"id":"dada55a8_67c84123","line":159,"in_reply_to":"dada55a8_0ced2ab7","updated":"2016-07-26 09:33:59.000000000","message":"I wanted to do a container restart after the mariadb upgrade to ensure that post-upgrade, everything is still happy. Yes, the upgrade plays control binary restarts, but performing a full container (and therfore along with it a mariadb restart) restart is test of both the new container config and mariadb version combined.\nThat was the reasoning at least.","commit_id":"5ad9631366b13d782f1c784e0d24fa103c206df0"}],"scripts/upgrade-utilities/playbooks/galera-cluster-rolling-restart.yml":[{"author":{"_account_id":7219,"name":"git-harry","email":"git-harry@live.co.uk","username":"git-harry"},"change_message_id":"cbd179709e00d213d3a632dd6ede2f869f36d5b6","unresolved":false,"context_lines":[{"line_number":16,"context_line":"- name: Gracefully restart mariadb/galera cluster"},{"line_number":17,"context_line":"  hosts: galera_all"},{"line_number":18,"context_line":"  serial: 1"},{"line_number":19,"context_line":"  max_fail_percentage: 100"},{"line_number":20,"context_line":"  gather_facts: false"},{"line_number":21,"context_line":"  user: root"},{"line_number":22,"context_line":"  tasks:"}],"source_content_type":"text/x-yaml","patch_set":1,"id":"dada55a8_0cfecaad","line":19,"updated":"2016-07-26 09:18:46.000000000","message":"According to the documentation [1] max_fail_percentage only causes a play to abort if its value is exceeded. Which would suggest that this play will never abort due to a failed host. I wonder if you meant to set it to 0, continuing to restart hosts that are part of a cluster after a failure strikes me as a bad idea if keeping the cluster active during the upgrade is the goal.\n\n[1] http://docs.ansible.com/ansible/playbooks_delegation.html#maximum-failure-percentage","commit_id":"5ad9631366b13d782f1c784e0d24fa103c206df0"},{"author":{"_account_id":425,"name":"Darren Birkett","email":"darren.birkett@gmail.com","username":"darren-birkett"},"change_message_id":"3b086b6fcab3b143e490b8cef7e97306629a9d68","unresolved":false,"context_lines":[{"line_number":16,"context_line":"- name: Gracefully restart mariadb/galera cluster"},{"line_number":17,"context_line":"  hosts: galera_all"},{"line_number":18,"context_line":"  serial: 1"},{"line_number":19,"context_line":"  max_fail_percentage: 100"},{"line_number":20,"context_line":"  gather_facts: false"},{"line_number":21,"context_line":"  user: root"},{"line_number":22,"context_line":"  tasks:"}],"source_content_type":"text/x-yaml","patch_set":1,"id":"dada55a8_07e06595","line":19,"in_reply_to":"dada55a8_0cfecaad","updated":"2016-07-26 09:33:59.000000000","message":"fixed in a later patchset","commit_id":"5ad9631366b13d782f1c784e0d24fa103c206df0"},{"author":{"_account_id":6816,"name":"Jesse Pretorius","email":"jesse@odyssey4.me","username":"jesse-pretorius"},"change_message_id":"860c9bff66d405aa47290646e47ef186c8022544","unresolved":false,"context_lines":[{"line_number":16,"context_line":"- name: Gracefully restart mariadb/galera cluster"},{"line_number":17,"context_line":"  hosts: galera_all"},{"line_number":18,"context_line":"  serial: 1"},{"line_number":19,"context_line":"  max_fail_percentage: 100"},{"line_number":20,"context_line":"  gather_facts: false"},{"line_number":21,"context_line":"  user: root"},{"line_number":22,"context_line":"  tasks:"}],"source_content_type":"text/x-yaml","patch_set":6,"id":"dada55a8_0ce08a95","line":19,"range":{"start_line":19,"start_character":2,"end_line":19,"end_character":26},"updated":"2016-07-26 09:16:15.000000000","message":"This is an odd entry. What\u0027s the intent here?","commit_id":"b74d73202835fef9ac9bad8650935223c8f8bd62"}]}