)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"d3c764acc30239781183a0decbe959773a264cc0","unresolved":false,"context_lines":[{"line_number":7,"context_line":"Correct default number of threads"},{"line_number":8,"context_line":""},{"line_number":9,"context_line":"The default number of threads does not match the value"},{"line_number":10,"context_line":"tested upstream, which can return errors for users when"},{"line_number":11,"context_line":"requests begin backing up in the webserver due to a lack"},{"line_number":12,"context_line":"of resources and the nature of the ironic API having some"},{"line_number":13,"context_line":"synchronous requests."}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"3f79a3b5_b5cd3a86","line":10,"updated":"2018-08-14 08:41:16.000000000","message":"Can we link somewhere to out CI configuration to back the claim?","commit_id":"efc454c304b04a10d14586e5a06a5c46e1178c03"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"a6c2524c3e80ca301cd43b22f4b8d193deda3634","unresolved":false,"context_lines":[{"line_number":7,"context_line":"Correct default number of threads"},{"line_number":8,"context_line":""},{"line_number":9,"context_line":"The default number of threads does not match the value"},{"line_number":10,"context_line":"tested upstream, which can return errors for users when"},{"line_number":11,"context_line":"requests begin backing up in the webserver due to a lack"},{"line_number":12,"context_line":"of resources and the nature of the ironic API having some"},{"line_number":13,"context_line":"synchronous requests."}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"3f79a3b5_24c4903f","line":10,"in_reply_to":"3f79a3b5_b5cd3a86","updated":"2018-08-14 20:12:44.000000000","message":"sure, I\u0027ll follow-up. :(","commit_id":"efc454c304b04a10d14586e5a06a5c46e1178c03"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"d3c764acc30239781183a0decbe959773a264cc0","unresolved":false,"context_lines":[{"line_number":10,"context_line":"tested upstream, which can return errors for users when"},{"line_number":11,"context_line":"requests begin backing up in the webserver due to a lack"},{"line_number":12,"context_line":"of resources and the nature of the ironic API having some"},{"line_number":13,"context_line":"synchronous requests."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"Change-Id: I327219f96cfc0a69643ab6369cb8e831251f799c"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"3f79a3b5_55c86672","line":13,"updated":"2018-08-14 08:41:16.000000000","message":"Please create a bug with the description of the errors we see.","commit_id":"efc454c304b04a10d14586e5a06a5c46e1178c03"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"a6c2524c3e80ca301cd43b22f4b8d193deda3634","unresolved":false,"context_lines":[{"line_number":10,"context_line":"tested upstream, which can return errors for users when"},{"line_number":11,"context_line":"requests begin backing up in the webserver due to a lack"},{"line_number":12,"context_line":"of resources and the nature of the ironic API having some"},{"line_number":13,"context_line":"synchronous requests."},{"line_number":14,"context_line":""},{"line_number":15,"context_line":"Change-Id: I327219f96cfc0a69643ab6369cb8e831251f799c"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"3f79a3b5_c4b35ce8","line":13,"in_reply_to":"3f79a3b5_55c86672","updated":"2018-08-14 20:12:44.000000000","message":"I can link to the rhbz, I see no purpose in duping bug tracking, especially for a something that generally improves the overall situation.","commit_id":"efc454c304b04a10d14586e5a06a5c46e1178c03"}],"manifests/wsgi/apache.pp":[{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"d3c764acc30239781183a0decbe959773a264cc0","unresolved":false,"context_lines":[{"line_number":50,"context_line":"#"},{"line_number":51,"context_line":"#   [*threads*]"},{"line_number":52,"context_line":"#     (optional) The number of threads for the vhost."},{"line_number":53,"context_line":"#     Defaults to 10"},{"line_number":54,"context_line":"#"},{"line_number":55,"context_line":"#   [*wsgi_process_display_name*]"},{"line_number":56,"context_line":"#     (optional) Name of the WSGI process display-name."}],"source_content_type":"text/x-puppet","patch_set":1,"id":"3f79a3b5_75c32258","line":53,"updated":"2018-08-14 08:41:16.000000000","message":"I wonder how 10 was picked.. Why not 8 or 16? Just thinking aloud.","commit_id":"efc454c304b04a10d14586e5a06a5c46e1178c03"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"a6c2524c3e80ca301cd43b22f4b8d193deda3634","unresolved":false,"context_lines":[{"line_number":50,"context_line":"#"},{"line_number":51,"context_line":"#   [*threads*]"},{"line_number":52,"context_line":"#     (optional) The number of threads for the vhost."},{"line_number":53,"context_line":"#     Defaults to 10"},{"line_number":54,"context_line":"#"},{"line_number":55,"context_line":"#   [*wsgi_process_display_name*]"},{"line_number":56,"context_line":"#     (optional) Name of the WSGI process display-name."}],"source_content_type":"text/x-puppet","patch_set":1,"id":"3f79a3b5_2476d03f","line":53,"in_reply_to":"3f79a3b5_75c32258","updated":"2018-08-14 20:12:44.000000000","message":"for devstack, no clue....\n\nI believe the default is like 12 for uwsgi.","commit_id":"efc454c304b04a10d14586e5a06a5c46e1178c03"},{"author":{"_account_id":10873,"name":"Juan Antonio Osorio Robles","email":"jaosorior@redhat.com","username":"ejuaoso"},"change_message_id":"139dbe7025fde522578eb851c2d304eb1fe38725","unresolved":false,"context_lines":[{"line_number":112,"context_line":"  $ssl_crl                     \u003d undef,"},{"line_number":113,"context_line":"  $ssl_certs_dir               \u003d undef,"},{"line_number":114,"context_line":"  $wsgi_process_display_name   \u003d undef,"},{"line_number":115,"context_line":"  $threads                     \u003d 15,"},{"line_number":116,"context_line":"  $priority                    \u003d \u002710\u0027,"},{"line_number":117,"context_line":"  $access_log_file             \u003d false,"},{"line_number":118,"context_line":"  $access_log_format           \u003d false,"}],"source_content_type":"text/x-puppet","patch_set":2,"id":"3f79a3b5_f72328bd","line":115,"updated":"2018-08-21 04:41:25.000000000","message":"In the past folks have complained about this kind of setting resulting in performance degradation due to Python\u0027s GIL. has this been tested to improve performance? If so, any idea why this would be the case? Would it be cause we\u0027re now running on python3?","commit_id":"037cd2cb5d332ba178d78390a3f5bcd75ce1e9e3"},{"author":{"_account_id":14985,"name":"Alex Schultz","email":"aschultz@next-development.com","username":"mwhahaha"},"change_message_id":"b33ae5a9ffe18f280f11b0d98f34652a509adca7","unresolved":false,"context_lines":[{"line_number":112,"context_line":"  $ssl_crl                     \u003d undef,"},{"line_number":113,"context_line":"  $ssl_certs_dir               \u003d undef,"},{"line_number":114,"context_line":"  $wsgi_process_display_name   \u003d undef,"},{"line_number":115,"context_line":"  $threads                     \u003d 15,"},{"line_number":116,"context_line":"  $priority                    \u003d \u002710\u0027,"},{"line_number":117,"context_line":"  $access_log_file             \u003d false,"},{"line_number":118,"context_line":"  $access_log_format           \u003d false,"}],"source_content_type":"text/x-puppet","patch_set":2,"id":"3f79a3b5_c6e0559a","line":115,"in_reply_to":"3f79a3b5_47657944","updated":"2018-08-24 13:50:08.000000000","message":"Unfortunately I think this is really service specific. Where in Keystone this is a bad, ironic it might be OK and preferred.","commit_id":"037cd2cb5d332ba178d78390a3f5bcd75ce1e9e3"},{"author":{"_account_id":11655,"name":"Julia Kreger","email":"juliaashleykreger@gmail.com","username":"jkreger","status":"Flying to the moon with a Jetpack!"},"change_message_id":"2cb6a049afab2c29ab8daa8fe149880b3503af80","unresolved":false,"context_lines":[{"line_number":112,"context_line":"  $ssl_crl                     \u003d undef,"},{"line_number":113,"context_line":"  $ssl_certs_dir               \u003d undef,"},{"line_number":114,"context_line":"  $wsgi_process_display_name   \u003d undef,"},{"line_number":115,"context_line":"  $threads                     \u003d 15,"},{"line_number":116,"context_line":"  $priority                    \u003d \u002710\u0027,"},{"line_number":117,"context_line":"  $access_log_file             \u003d false,"},{"line_number":118,"context_line":"  $access_log_format           \u003d false,"}],"source_content_type":"text/x-puppet","patch_set":2,"id":"3f79a3b5_47657944","line":115,"in_reply_to":"3f79a3b5_f72328bd","updated":"2018-08-21 19:00:43.000000000","message":"I believe this was discovered on python2.7, but likely applies to python3 as well as the issue is how we\u0027re working with apache. Specifically the issue was discovered with high concurrency, except with a limit of 2 or 4 processes results in requests stacking up. Given ironic has synchronous \n some requests, the user was experience requests stacking up in apache and eventually failing.\n\nChanging it to include threads increased the possible concurrency, which prevented the requests from backing up in apache.\n\nThis was not explicitly performance tested, but the user compared a newton deployment with an ironic-api process which is multithreaded by default versus a queens deployment where generated by the prior configuration of 4 processes and 1 thread. The users testing with high concurrency and as far as I\u0027m aware no noticeable degradation in performance for their requests.  Given the nature of the API, I think concurrency is far more desirable that anything else.","commit_id":"037cd2cb5d332ba178d78390a3f5bcd75ce1e9e3"}]}