)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":13252,"name":"Dr. Jens Harbott","display_name":"Jens Harbott (frickler)","email":"frickler@offenerstapel.de","username":"jrosenboom"},"change_message_id":"432b314f9f59283392ef5db312c4c6af5f8c7303","unresolved":true,"context_lines":[{"line_number":9,"context_line":"Set writer_is_also_reader to 2 in the Galera hostgroup"},{"line_number":10,"context_line":"configuration so that reads are distributed across all"},{"line_number":11,"context_line":"available backend nodes instead of being routed only to"},{"line_number":12,"context_line":"the writer. With writer_is_also_reader set to 0, all read"},{"line_number":13,"context_line":"traffic went to a single node, causing unnecessary load and"},{"line_number":14,"context_line":"higher latency. See bug report for performance difference"},{"line_number":15,"context_line":"details."}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"5946afeb_56b9ae09","line":12,"range":{"start_line":12,"start_character":4,"end_line":12,"end_character":10},"updated":"2026-09-11 08:50:03.000000000","message":"I\u0027m not sure that this interpretation is correct, looking at https://www.proxysql.com/documentation/galera-configuration/ the writer should never be in the reader_hostgroup, but the instead the backup_writer was also not being used for reads, leaving only the single third server for reads. This setup might be fine for deployments that have more than 3 database nodes, but I think the change is useful for most deployments anyway, just let\u0027s use the correct description here and in the reno","commit_id":"44b8f620b0230d84dbb7397aa558da1d4c840980"},{"author":{"_account_id":34034,"name":"Franciszek Przewoźny","display_name":"Franciszek Przewozny","email":"przewozny.franciszek@gmail.com","username":"fprzewozn","status":"OpenStack Engineer / SRE @ Opera Software"},"change_message_id":"26144bcef4dcf167cb20528d542008a13bbdd436","unresolved":false,"context_lines":[{"line_number":9,"context_line":"Set writer_is_also_reader to 2 in the Galera hostgroup"},{"line_number":10,"context_line":"configuration so that reads are distributed across all"},{"line_number":11,"context_line":"available backend nodes instead of being routed only to"},{"line_number":12,"context_line":"the writer. With writer_is_also_reader set to 0, all read"},{"line_number":13,"context_line":"traffic went to a single node, causing unnecessary load and"},{"line_number":14,"context_line":"higher latency. See bug report for performance difference"},{"line_number":15,"context_line":"details."}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"00e364fa_f200b92b","line":12,"range":{"start_line":12,"start_character":4,"end_line":12,"end_character":10},"in_reply_to":"5946afeb_56b9ae09","updated":"2026-09-11 12:07:20.000000000","message":"Done","commit_id":"44b8f620b0230d84dbb7397aa558da1d4c840980"}],"/PATCHSET_LEVEL":[{"author":{"_account_id":34034,"name":"Franciszek Przewoźny","display_name":"Franciszek Przewozny","email":"przewozny.franciszek@gmail.com","username":"fprzewozn","status":"OpenStack Engineer / SRE @ Opera Software"},"change_message_id":"0bae83a7c27336e087b90c36656c4e019d681c28","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"f13bbd38_2181b05d","updated":"2026-09-11 05:58:20.000000000","message":"recheck","commit_id":"44b8f620b0230d84dbb7397aa558da1d4c840980"},{"author":{"_account_id":37203,"name":"Bertrand Lanson","display_name":"Bertrand Lanson","email":"bertrand.lanson@infomaniak.com","username":"lanson","status":"Infomaniak Network SA"},"change_message_id":"a5eddfe73c438926017278468f0087a13de192d0","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"27cff15b_d6342476","updated":"2026-09-15 13:36:40.000000000","message":"LGTM","commit_id":"42eec877ff572fd724af8a1c0fa64acf5e824c7b"},{"author":{"_account_id":22629,"name":"Michal Nasiadka","email":"mnasiadka@gmail.com","username":"mnasiadka"},"change_message_id":"5f25bb2ea09002cd32dab343f06167969dc94cfb","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"8a977658_68577788","updated":"2026-09-16 13:13:31.000000000","message":"Makes sense given:\n\nFlag with three possible values: (0, 1, 2).\n- writer_is_also_reader\u003d0: nodes in the writer_hostgroup and in the backup_writer_hostgroup wont be in reader hostgroup.\n- writer_is_also_reader\u003d1: nodes in writer_hostgroup and backup_writer_hostgroup are also in reader hostgroup.\n- writer_is_also_reader\u003d2 : nodes in backup_writer_hostgroup are also in reader_hostgroup.\n\nWe can also define the action taken for the nodes in terms of the READ_ONLY flag:\n- writer_is_also_reader\u003d0: Nodes with READ_ONLY\u003d0 are removed for reader_hostgroup placed in hostgroups backup_writer_hostgroup and writer_hostgroup as corresponds.\n- writer_is_also_reader\u003d1: The nodes in the writer_hostgroup and in the backup_writer_hostgroup are also in the reader_hostgroup.\n- writer_is_also_reader\u003d2: The nodes in the backup_writer_hostgroup are also in the reader hostgroup.","commit_id":"42eec877ff572fd724af8a1c0fa64acf5e824c7b"},{"author":{"_account_id":34034,"name":"Franciszek Przewoźny","display_name":"Franciszek Przewozny","email":"przewozny.franciszek@gmail.com","username":"fprzewozn","status":"OpenStack Engineer / SRE @ Opera Software"},"change_message_id":"8cfcd30f9d2feef51ab474c2fa19809b71d2e68d","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"bbff110e_2942b4b0","updated":"2026-09-11 18:35:12.000000000","message":"recheck","commit_id":"42eec877ff572fd724af8a1c0fa64acf5e824c7b"}]}