)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":30491,"name":"Radosław Piliszek","display_name":"Radek","email":"radek@piliszek.it","username":"yoctozepto","status":"self-employed techologist, collaborating mostly with 7bulls.com"},"change_message_id":"c27524b441b5d4d089846d1e06535c05d6e03221","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"77b8c23c_a3fb7b24","updated":"2022-04-21 17:52:55.000000000","message":"tbh, I still think we don\u0027t need it - as the VIP address is on the \"master\" node and thus that\u0027s why there was a split there already - the backups are fine to restart quickly; please elaborate further; also, this patch reintroduces the bug we had previously - that there are hanging connections being kept because of the lost VIP address","commit_id":"4666acd8f26d2101b209951d665ca22e3a6cef54"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"88ed718029b6b4952ae946e3724c505881b80e68","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":2,"id":"8915a8ee_b15d5043","in_reply_to":"53a85ee3_788ec323","updated":"2022-04-28 11:31:00.000000000","message":"Any reply ? 😕","commit_id":"4666acd8f26d2101b209951d665ca22e3a6cef54"},{"author":{"_account_id":30491,"name":"Radosław Piliszek","display_name":"Radek","email":"radek@piliszek.it","username":"yoctozepto","status":"self-employed techologist, collaborating mostly with 7bulls.com"},"change_message_id":"69cca1cb0c9edd5fa0a4945b3d7016472f10341f","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":2,"id":"969e4cd8_587ff6c0","in_reply_to":"6edf87db_adccd016","updated":"2022-04-21 18:20:51.000000000","message":"OK, let\u0027s split this thread in two.\n\n1) There is no VIP on backups so there is nothing to trigger in keepalived - this is effectively a slower noop.\n\n2) Yours is not default and helps in the general case (e.g., when the node dies). This trick was to let kernel send TCP RST for haproxy connections (thus your wait is irrelevant unless the TCP RST were dropped in the network). I guess we could rely on your and forget about this one but I don\u0027t see why we would want to change the order at all.\n\nPS: That\u0027s why I wanted you to refactor this change, I felt it does something weird.","commit_id":"4666acd8f26d2101b209951d665ca22e3a6cef54"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"403aef31b1634884f7a39e4083da7cb737b6e742","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"6edf87db_adccd016","in_reply_to":"77b8c23c_a3fb7b24","updated":"2022-04-21 17:58:01.000000000","message":"Reintroduces by which part ? \nBecause if i remember hanging connections was fixed by me by ipv4.net.tcp.retries2 parameter in sysctl...\n\nAnd I really  remember that I needed to split that logic.","commit_id":"4666acd8f26d2101b209951d665ca22e3a6cef54"},{"author":{"_account_id":30491,"name":"Radosław Piliszek","display_name":"Radek","email":"radek@piliszek.it","username":"yoctozepto","status":"self-employed techologist, collaborating mostly with 7bulls.com"},"change_message_id":"654c8ce1991e2345f8186cc43a249d1e344f82d7","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"7d13fc1a_3eaf347e","in_reply_to":"8915a8ee_b15d5043","updated":"2022-05-02 10:57:16.000000000","message":"I amended the code as I understood what you needed it for. I will rebase the other ones and have further inquiries on them. This one is gtg as far as I am concerned.","commit_id":"4666acd8f26d2101b209951d665ca22e3a6cef54"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"fc20be819048d885593eb8cda39b23a059322fd9","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":2,"id":"53a85ee3_788ec323","in_reply_to":"8c0184c6_32148436","updated":"2022-04-22 12:03:02.000000000","message":"What I\u0027m trying to say that I didn\u0027t change anything, just split handler Restart to stop and start. And it\u0027s because ansible is calling handlers serially.","commit_id":"4666acd8f26d2101b209951d665ca22e3a6cef54"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"baea96b03ca8245d1ae5babe6d7220452cbfe286","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":2,"id":"8c0184c6_32148436","in_reply_to":"969e4cd8_587ff6c0","updated":"2022-04-21 19:47:35.000000000","message":"1 \n--------\n\n\nHmm, I am not sure If i remember correctly, but maybe the reason why I splitted this restart was switching balancer for mariadb (enable_proxysql: yes/no)\n\n1. haproxy -\u003e proxysql \\\n                           different order\n2. Proxysql -\u003e haproxy  /\n\nThant means, handlers are always called in some order for example in case:\n\n- restart haproxy\n- restart proxysql\n\n1. This if fine, because services.d/mariadb is removed so haproxy is restarted and then proxysql is started with listening on 3306.\n\n2. Proxysql is listening 3306 but ansible is calling handlers in order defined above, so services.d/mariadb is defined ..haproxy is restarted, but this will fail  because proxysql is not restarted and it is just going to be restarted in second handler call, but this of course - will fail.\n\nWhen it is \n\n- Stop haproxy\n- Stop proxysql\n- Start haproxy\n- Start proxysql\n\nEverything is ok, because in both cases both containers is stoppped and then started, so no conflicting port.\n\nYeah, it is slower, but it is not big deal, because this action is done on all nodes expect current master node, and then when it is running on node which is current master VIP is switched in moment when one container is down , because of keepalived check, this solution we are using from stein on production ..and never had any issue.","commit_id":"4666acd8f26d2101b209951d665ca22e3a6cef54"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"095c236d14f350c7f0b48826ee618671801949ae","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":3,"id":"a875cb81_cab4ec30","updated":"2022-05-02 10:57:17.000000000","message":"Thanks :)","commit_id":"9b91c3b896c96524ce929d3a069b0c1bc9e46b8e"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"daf8ffcbfca21af228377fac58203b4f64082d3b","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":5,"id":"a98bd2ca_c94daa55","updated":"2022-05-04 10:00:50.000000000","message":"Previous patch from Radoslaw was working but when I tested various scenarios in my test environment there were cases when it was not working as expected, as previous patch was working always, returning to that one and if there will be changes ..let\u0027s discuss and change it step by step by small pieces.","commit_id":"c4606a350abeea5bba10264cec1337a493846778"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"611ebaa3222fcf130475479e796f5bb0a4fac037","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":5,"id":"6e9d2915_acac5e68","in_reply_to":"07a24560_fddcf917","updated":"2022-05-04 10:12:45.000000000","message":"Or maybe i had corrupted my stack, so let me test it again :). It took me a while to figure out why those cells didn\u0027t work ..but then I understood.. So let\u0027s wait for first iteration of zuul tests.","commit_id":"c4606a350abeea5bba10264cec1337a493846778"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"204abcb931a67f728601b7062917cc8ed67716bc","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":5,"id":"ab6e9e67_82046f2f","in_reply_to":"6e9d2915_acac5e68","updated":"2022-05-05 10:33:18.000000000","message":"When config is changed for  haproxy, and VIP is switched to backup ..previous master should be restarted (I think not only haproxy, but also proxysql)\n\nYour patch did this : \n\nRUNNING HANDLER [loadbalancer : Check IP addresses on the API interface] ********************************************************************************************************************************************************************\nok: [controller0]\nok: [controller1]\n\nRUNNING HANDLER [loadbalancer : Group HA nodes by status] ***********************************************************************************************************************************************************************************\nok: [controller0]\nok: [controller1]\n\nRUNNING HANDLER [loadbalancer : Stop backup haproxy container] ******************************************************************************************************************************************************************************\nskipping: [controller0]\nchanged: [controller1]\n\nRUNNING HANDLER [loadbalancer : Start backup haproxy container] *****************************************************************************************************************************************************************************\nskipping: [controller0]\nchanged: [controller1]\n\nRUNNING HANDLER [loadbalancer : Wait for backup haproxy to start] ***************************************************************************************************************************************************************************\nok: [controller1]\n\nRUNNING HANDLER [loadbalancer : Stop master haproxy container] ******************************************************************************************************************************************************************************\nskipping: [controller1]\nchanged: [controller0]\n\nRUNNING HANDLER [loadbalancer : Start master haproxy container] *****************************************************************************************************************************************************************************\nskipping: [controller1]\nchanged: [controller0]\n\nRUNNING HANDLER [loadbalancer : Wait for master haproxy to start] ***************************************************************************************************************************************************************************\nok: [controller0]\n\nI think my patch worked different and also restarted proxysql, but let me check... for now I fix another patchset from chain as I found a bug in cells ..so now it will pass I hope.","commit_id":"c4606a350abeea5bba10264cec1337a493846778"},{"author":{"_account_id":30491,"name":"Radosław Piliszek","display_name":"Radek","email":"radek@piliszek.it","username":"yoctozepto","status":"self-employed techologist, collaborating mostly with 7bulls.com"},"change_message_id":"b4f1d7b9106615ebd9f927c523435945338ce0d5","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":5,"id":"ba716b7d_56dd1313","in_reply_to":"a98bd2ca_c94daa55","updated":"2022-05-04 10:01:55.000000000","message":"What was not working? Please be more specific.","commit_id":"c4606a350abeea5bba10264cec1337a493846778"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"8c1c0f21be31392d03b27da93f06080ed58f70e3","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":5,"id":"07a24560_fddcf917","in_reply_to":"ba716b7d_56dd1313","updated":"2022-05-04 10:03:49.000000000","message":"I\u0027m going to test it and will paste it here, for now I mainly want to see the result of CI test for cells.","commit_id":"c4606a350abeea5bba10264cec1337a493846778"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"5a5ea54864b22f4195e215e59951d3a9a12e142f","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":6,"id":"0933fcc8_02bd7b79","updated":"2022-06-02 10:26:45.000000000","message":"I think this can be merged now as is, i will wait for a while and then I will give +W.","commit_id":"959b7f79212cf8fee7f8c0d7abb124b37270798e"},{"author":{"_account_id":30491,"name":"Radosław Piliszek","display_name":"Radek","email":"radek@piliszek.it","username":"yoctozepto","status":"self-employed techologist, collaborating mostly with 7bulls.com"},"change_message_id":"5ce8f468dc93369ce466fc320ddb4346f15688fb","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":6,"id":"6c4ca0db_83227102","updated":"2022-06-02 10:27:44.000000000","message":"it\u0027s ok, let it merge normally","commit_id":"959b7f79212cf8fee7f8c0d7abb124b37270798e"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"e984457178ab16fc37138e01f27c4a9b4fb680a0","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":6,"id":"f96e7596_24311c70","in_reply_to":"0933fcc8_02bd7b79","updated":"2022-06-02 10:27:14.000000000","message":"Ups, i\u0027ve already clicked them :D","commit_id":"959b7f79212cf8fee7f8c0d7abb124b37270798e"},{"author":{"_account_id":30491,"name":"Radosław Piliszek","display_name":"Radek","email":"radek@piliszek.it","username":"yoctozepto","status":"self-employed techologist, collaborating mostly with 7bulls.com"},"change_message_id":"9f94f1337bb30c034d0cfe26c7411e7ec3051f64","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":6,"id":"6c0ae088_5ccafff8","in_reply_to":"0933fcc8_02bd7b79","updated":"2022-06-02 10:27:26.000000000","message":"You gave it already.","commit_id":"959b7f79212cf8fee7f8c0d7abb124b37270798e"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"ebf89ef6db7501d1c4072d20ba8f6d5eb8a9b568","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":6,"id":"ca7ef71b_84b2991d","in_reply_to":"6c4ca0db_83227102","updated":"2022-06-02 10:32:31.000000000","message":"Check follow-up patch, i\u0027ve pasted output from real environment there.","commit_id":"959b7f79212cf8fee7f8c0d7abb124b37270798e"}],"ansible/roles/loadbalancer/handlers/main.yml":[{"author":{"_account_id":30491,"name":"Radosław Piliszek","display_name":"Radek","email":"radek@piliszek.it","username":"yoctozepto","status":"self-employed techologist, collaborating mostly with 7bulls.com"},"change_message_id":"c27524b441b5d4d089846d1e06535c05d6e03221","unresolved":true,"context_lines":[{"line_number":50,"context_line":"  become: true"},{"line_number":51,"context_line":"  kolla_docker:"},{"line_number":52,"context_line":"    action: \"stop_container\""},{"line_number":53,"context_line":"    # NOTE(kevko): backup node might not have keepalived yet - ignore"},{"line_number":54,"context_line":"    ignore_missing: true"},{"line_number":55,"context_line":"    common_options: \"{{ docker_common_options }}\""},{"line_number":56,"context_line":"    name: \"haproxy\""}],"source_content_type":"text/x-yaml","patch_set":2,"id":"3b86ed98_93fc239b","line":53,"range":{"start_line":53,"start_character":46,"end_line":53,"end_character":56},"updated":"2022-04-21 17:52:55.000000000","message":"haproxy","commit_id":"4666acd8f26d2101b209951d665ca22e3a6cef54"},{"author":{"_account_id":30491,"name":"Radosław Piliszek","display_name":"Radek","email":"radek@piliszek.it","username":"yoctozepto","status":"self-employed techologist, collaborating mostly with 7bulls.com"},"change_message_id":"654c8ce1991e2345f8186cc43a249d1e344f82d7","unresolved":false,"context_lines":[{"line_number":50,"context_line":"  become: true"},{"line_number":51,"context_line":"  kolla_docker:"},{"line_number":52,"context_line":"    action: \"stop_container\""},{"line_number":53,"context_line":"    # NOTE(kevko): backup node might not have keepalived yet - ignore"},{"line_number":54,"context_line":"    ignore_missing: true"},{"line_number":55,"context_line":"    common_options: \"{{ docker_common_options }}\""},{"line_number":56,"context_line":"    name: \"haproxy\""}],"source_content_type":"text/x-yaml","patch_set":2,"id":"ebb67c16_d27be1f0","line":53,"range":{"start_line":53,"start_character":46,"end_line":53,"end_character":56},"in_reply_to":"3b86ed98_93fc239b","updated":"2022-05-02 10:57:16.000000000","message":"Done","commit_id":"4666acd8f26d2101b209951d665ca22e3a6cef54"},{"author":{"_account_id":30491,"name":"Radosław Piliszek","display_name":"Radek","email":"radek@piliszek.it","username":"yoctozepto","status":"self-employed techologist, collaborating mostly with 7bulls.com"},"change_message_id":"c27524b441b5d4d089846d1e06535c05d6e03221","unresolved":true,"context_lines":[{"line_number":115,"context_line":"  notify:"},{"line_number":116,"context_line":"    - Wait for virtual IP to appear"},{"line_number":117,"context_line":""},{"line_number":118,"context_line":"- name: Stop master keepalived container"},{"line_number":119,"context_line":"  become: true"},{"line_number":120,"context_line":"  kolla_docker:"},{"line_number":121,"context_line":"    action: \"stop_container\""}],"source_content_type":"text/x-yaml","patch_set":2,"id":"f9b9a692_96cfa093","line":118,"updated":"2022-04-21 17:52:55.000000000","message":"switching order here will mean that we will have connections that were not dropped","commit_id":"4666acd8f26d2101b209951d665ca22e3a6cef54"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"baea96b03ca8245d1ae5babe6d7220452cbfe286","unresolved":true,"context_lines":[{"line_number":115,"context_line":"  notify:"},{"line_number":116,"context_line":"    - Wait for virtual IP to appear"},{"line_number":117,"context_line":""},{"line_number":118,"context_line":"- name: Stop master keepalived container"},{"line_number":119,"context_line":"  become: true"},{"line_number":120,"context_line":"  kolla_docker:"},{"line_number":121,"context_line":"    action: \"stop_container\""}],"source_content_type":"text/x-yaml","patch_set":2,"id":"ef5164da_f599c266","line":118,"in_reply_to":"7bb91ee2_be8f3929","updated":"2022-04-21 19:47:35.000000000","message":"yes it is, point 1 i tried to explain above.","commit_id":"4666acd8f26d2101b209951d665ca22e3a6cef54"},{"author":{"_account_id":30491,"name":"Radosław Piliszek","display_name":"Radek","email":"radek@piliszek.it","username":"yoctozepto","status":"self-employed techologist, collaborating mostly with 7bulls.com"},"change_message_id":"69cca1cb0c9edd5fa0a4945b3d7016472f10341f","unresolved":true,"context_lines":[{"line_number":115,"context_line":"  notify:"},{"line_number":116,"context_line":"    - Wait for virtual IP to appear"},{"line_number":117,"context_line":""},{"line_number":118,"context_line":"- name: Stop master keepalived container"},{"line_number":119,"context_line":"  become: true"},{"line_number":120,"context_line":"  kolla_docker:"},{"line_number":121,"context_line":"    action: \"stop_container\""}],"source_content_type":"text/x-yaml","patch_set":2,"id":"7bb91ee2_be8f3929","line":118,"in_reply_to":"d0bcfd29_932eeb8a","updated":"2022-04-21 18:20:51.000000000","message":"mhm; this is point 2 from the main comment thread here","commit_id":"4666acd8f26d2101b209951d665ca22e3a6cef54"},{"author":{"_account_id":30491,"name":"Radosław Piliszek","display_name":"Radek","email":"radek@piliszek.it","username":"yoctozepto","status":"self-employed techologist, collaborating mostly with 7bulls.com"},"change_message_id":"654c8ce1991e2345f8186cc43a249d1e344f82d7","unresolved":false,"context_lines":[{"line_number":115,"context_line":"  notify:"},{"line_number":116,"context_line":"    - Wait for virtual IP to appear"},{"line_number":117,"context_line":""},{"line_number":118,"context_line":"- name: Stop master keepalived container"},{"line_number":119,"context_line":"  become: true"},{"line_number":120,"context_line":"  kolla_docker:"},{"line_number":121,"context_line":"    action: \"stop_container\""}],"source_content_type":"text/x-yaml","patch_set":2,"id":"03a4df47_98fa1043","line":118,"in_reply_to":"ef5164da_f599c266","updated":"2022-05-02 10:57:16.000000000","message":"Done","commit_id":"4666acd8f26d2101b209951d665ca22e3a6cef54"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"403aef31b1634884f7a39e4083da7cb737b6e742","unresolved":true,"context_lines":[{"line_number":115,"context_line":"  notify:"},{"line_number":116,"context_line":"    - Wait for virtual IP to appear"},{"line_number":117,"context_line":""},{"line_number":118,"context_line":"- name: Stop master keepalived container"},{"line_number":119,"context_line":"  become: true"},{"line_number":120,"context_line":"  kolla_docker:"},{"line_number":121,"context_line":"    action: \"stop_container\""}],"source_content_type":"text/x-yaml","patch_set":2,"id":"d0bcfd29_932eeb8a","line":118,"in_reply_to":"f9b9a692_96cfa093","updated":"2022-04-21 17:58:01.000000000","message":"# NOTE(yoctozepto): This is to ensure haproxy can close any open connections\n# to the VIP address. \n\nline 133 ?","commit_id":"4666acd8f26d2101b209951d665ca22e3a6cef54"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"08278d15685b3badb4abe2d04af1353ad3c31183","unresolved":true,"context_lines":[{"line_number":133,"context_line":"    - groups.kolla_ha_is_master_True is defined"},{"line_number":134,"context_line":"    - inventory_hostname in groups.kolla_ha_is_master_True"},{"line_number":135,"context_line":"  listen:"},{"line_number":136,"context_line":"    - Restart haproxy container"},{"line_number":137,"context_line":"    - Restart keepalived container"},{"line_number":138,"context_line":""},{"line_number":139,"context_line":"- name: Stop master keepalived container"}],"source_content_type":"text/x-yaml","patch_set":3,"id":"0159d5fc_0775290b","line":136,"range":{"start_line":136,"start_character":0,"end_line":136,"end_character":31},"updated":"2022-05-31 08:50:37.000000000","message":"Hmm, it looks like you are fixing buggy code which is currently in master and which has been overlooked so far. Am I right ? \n\nWhen I analalyzed code in detail.\n\n\nCurrent situation: \n\nIn case notify \"Restart keepalived\" :\n\n1.  Check IP addresses on the API interface\n2.  Group HA nodes by status\n3.  Stop backup keepalived container\n4.  Restart backup haproxy container\n5.  Wait for backup haproxy to start\n6.  Start backup keepalived container\n7.  Wait for virtual IP to appear\n8.  Stop master haproxy container\n9.  Stop master keepalived container\n10. Start master haproxy container\n11. Wait for master haproxy to start\n12. Start master keepalived container\n13. Wait for virtual IP to appear\n\n\nIn case notify \"Restart haproxy\" :\n\n1.  Check IP addresses on the API interface\n2.  Group HA nodes by status\n3.  Restart backup haproxy container\n4.  Wait for backup haproxy to start\n5.  Start master haproxy container\n6.  Wait for master haproxy to start\n\nYour proposal change patch 3:\n\n\nIn case notify \"Restart keepalived\" :\n\n1.  Check IP addresses on the API interface\n2.  Group HA nodes by status\n3.  Stop backup keepalived container\n4.  Stop backup haproxy container      \\  Original \"Restart backup haproxy container\"\n5.  Start backup haproxy container     /\n6.  Wait for backup haproxy to start\n7.  Start backup keepalived container\n8.  Wait for virtual IP to appear\n9.  Stop master haproxy container\n10. Stop master keepalived container\n11. Start master haproxy container\n12. Wait for master haproxy to start\n13. Start master keepalived container\n14. Wait for virtual IP to appear\n\nIn case notify \"Restart haproxy\" :\n\n1.  Check IP addresses on the API interface\n2.  Group HA nodes by status\n3.  Stop backup haproxy container       \\ Original \"Restart backup haproxy container\"\n4.  Start backup haproxy container      / \n5.  Wait for backup haproxy to start\n6.  Stop master haproxy container      - This is one task more - which looks like it\u0027s correct and master code is broken\n7.  Start master haproxy container\n8.  Wait for master haproxy to start","commit_id":"9b91c3b896c96524ce929d3a069b0c1bc9e46b8e"},{"author":{"_account_id":30491,"name":"Radosław Piliszek","display_name":"Radek","email":"radek@piliszek.it","username":"yoctozepto","status":"self-employed techologist, collaborating mostly with 7bulls.com"},"change_message_id":"f35ce84453fca99b6116f9238352d5b84c02922a","unresolved":false,"context_lines":[{"line_number":133,"context_line":"    - groups.kolla_ha_is_master_True is defined"},{"line_number":134,"context_line":"    - inventory_hostname in groups.kolla_ha_is_master_True"},{"line_number":135,"context_line":"  listen:"},{"line_number":136,"context_line":"    - Restart haproxy container"},{"line_number":137,"context_line":"    - Restart keepalived container"},{"line_number":138,"context_line":""},{"line_number":139,"context_line":"- name: Stop master keepalived container"}],"source_content_type":"text/x-yaml","patch_set":3,"id":"ac8999c7_e77c2fd6","line":136,"range":{"start_line":136,"start_character":0,"end_line":136,"end_character":31},"in_reply_to":"0159d5fc_0775290b","updated":"2022-05-31 09:07:41.000000000","message":"No, we don\u0027t need to stop haproxy when keepalived is not restarted. We should remove this listen. It got here by mistake (mind I have been refactoring your code).","commit_id":"9b91c3b896c96524ce929d3a069b0c1bc9e46b8e"},{"author":{"_account_id":27339,"name":"Michal Arbet","email":"michal.arbet@ultimum.io","username":"michalarbet"},"change_message_id":"9366749deace14b6d5b628b827d512b1d38419d8","unresolved":false,"context_lines":[{"line_number":133,"context_line":"    - groups.kolla_ha_is_master_True is defined"},{"line_number":134,"context_line":"    - inventory_hostname in groups.kolla_ha_is_master_True"},{"line_number":135,"context_line":"  listen:"},{"line_number":136,"context_line":"    - Restart haproxy container"},{"line_number":137,"context_line":"    - Restart keepalived container"},{"line_number":138,"context_line":""},{"line_number":139,"context_line":"- name: Stop master keepalived container"}],"source_content_type":"text/x-yaml","patch_set":3,"id":"f1a21d67_6e1cf0bf","line":136,"range":{"start_line":136,"start_character":0,"end_line":136,"end_character":31},"in_reply_to":"ac8999c7_e77c2fd6","updated":"2022-05-31 09:11:54.000000000","message":"Ok, then my original patch was correct as I am not touching this one.","commit_id":"9b91c3b896c96524ce929d3a069b0c1bc9e46b8e"},{"author":{"_account_id":30491,"name":"Radosław Piliszek","display_name":"Radek","email":"radek@piliszek.it","username":"yoctozepto","status":"self-employed techologist, collaborating mostly with 7bulls.com"},"change_message_id":"3b0442ed0c788fe2516affb33f1c5734699ee351","unresolved":false,"context_lines":[{"line_number":133,"context_line":"    - groups.kolla_ha_is_master_True is defined"},{"line_number":134,"context_line":"    - inventory_hostname in groups.kolla_ha_is_master_True"},{"line_number":135,"context_line":"  listen:"},{"line_number":136,"context_line":"    - Restart haproxy container"},{"line_number":137,"context_line":"    - Restart keepalived container"},{"line_number":138,"context_line":""},{"line_number":139,"context_line":"- name: Stop master keepalived container"}],"source_content_type":"text/x-yaml","patch_set":3,"id":"08594411_790080a3","line":136,"range":{"start_line":136,"start_character":0,"end_line":136,"end_character":31},"in_reply_to":"f1a21d67_6e1cf0bf","updated":"2022-05-31 09:18:09.000000000","message":"Ack, then bad refactoring of mine. :-)","commit_id":"9b91c3b896c96524ce929d3a069b0c1bc9e46b8e"}]}