)]}'
{"doc/source/devref/error_handling.rst":[{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"d9f067c00040ea3f6864ff4109e3b26c4dc9b008","unresolved":false,"context_lines":[{"line_number":18,"context_line":"patterns to recover from patterns."},{"line_number":19,"context_line":""},{"line_number":20,"context_line":"Then we look to the future and how this may be evolved into a more consistent"},{"line_number":21,"context_line":"approach."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"Logging"},{"line_number":24,"context_line":"--------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a0c5dc1_6d99566e","line":21,"updated":"2015-05-13 08:31:54.000000000","message":"Is the goal of this doc to explain what\u0027s there and to give insights for improving what we have today ? All my below comments are based on that.","commit_id":"a09ffbf34e078e9ef1b755130d23594d1cc62ebe"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"779625aae2b1eb0d0392b6afc87e62ceca4236b2","unresolved":false,"context_lines":[{"line_number":18,"context_line":"patterns to recover from patterns."},{"line_number":19,"context_line":""},{"line_number":20,"context_line":"Then we look to the future and how this may be evolved into a more consistent"},{"line_number":21,"context_line":"approach."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"Logging"},{"line_number":24,"context_line":"--------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a0c5dc1_b8fae5bb","line":21,"in_reply_to":"9a0c5dc1_6d99566e","updated":"2015-05-13 10:50:56.000000000","message":"yes","commit_id":"a09ffbf34e078e9ef1b755130d23594d1cc62ebe"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"d9f067c00040ea3f6864ff4109e3b26c4dc9b008","unresolved":false,"context_lines":[{"line_number":25,"context_line":""},{"line_number":26,"context_line":"We now have cross project guidelines on logging, and Nova embraces these plans"},{"line_number":27,"context_line":"stated here: todo"},{"line_number":28,"context_line":""},{"line_number":29,"context_line":"There is much work needed to fully align with the above. Key problems include"},{"line_number":30,"context_line":"having useful information in debug logs that forces operators to keep debug"},{"line_number":31,"context_line":"logs, having too many stack traces in the error logs that leads to very large"}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a0c5dc1_0d1e5a0c","line":28,"updated":"2015-05-13 08:31:54.000000000","message":"we can talk about the most known issues for debugging using logs : the scheduler is a good one.\n\nHaving also the layer isolation enforcement makes things a little hard since for example the scheduler doesn\u0027t know which instances he has, he just knows how many he has to decide.\n\nThe problem is also the tracability of the request, since the context request ID is not persisted, which makes it very hard to follow from the API to the conductor and then compute.\n\nThat\u0027s the few thoughts I just get now.","commit_id":"a09ffbf34e078e9ef1b755130d23594d1cc62ebe"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"779625aae2b1eb0d0392b6afc87e62ceca4236b2","unresolved":false,"context_lines":[{"line_number":25,"context_line":""},{"line_number":26,"context_line":"We now have cross project guidelines on logging, and Nova embraces these plans"},{"line_number":27,"context_line":"stated here: todo"},{"line_number":28,"context_line":""},{"line_number":29,"context_line":"There is much work needed to fully align with the above. Key problems include"},{"line_number":30,"context_line":"having useful information in debug logs that forces operators to keep debug"},{"line_number":31,"context_line":"logs, having too many stack traces in the error logs that leads to very large"}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a0c5dc1_f8ddfd42","line":28,"in_reply_to":"9a0c5dc1_0d1e5a0c","updated":"2015-05-13 10:50:56.000000000","message":"Not sure about all that. Thats not really all related to logs.","commit_id":"a09ffbf34e078e9ef1b755130d23594d1cc62ebe"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"d9f067c00040ea3f6864ff4109e3b26c4dc9b008","unresolved":false,"context_lines":[{"line_number":39,"context_line":"than most logs. These are not just about error handling, they are often used"},{"line_number":40,"context_line":"for billing. More worryingly they are unversioned and (largely) untested."},{"line_number":41,"context_line":"Every instance save sends a notification of its update. That can, with the"},{"line_number":42,"context_line":"help of stacktach, actually be used to profile every large step of the boot"},{"line_number":43,"context_line":"process."},{"line_number":44,"context_line":""},{"line_number":45,"context_line":"Instance Faults"}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a0c5dc1_4db9f2e8","line":42,"updated":"2015-05-13 08:31:54.000000000","message":"I would not mention stacktach directly, just saying \"with the help of a 3rd party tool eating the notifications\"","commit_id":"a09ffbf34e078e9ef1b755130d23594d1cc62ebe"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"779625aae2b1eb0d0392b6afc87e62ceca4236b2","unresolved":false,"context_lines":[{"line_number":39,"context_line":"than most logs. These are not just about error handling, they are often used"},{"line_number":40,"context_line":"for billing. More worryingly they are unversioned and (largely) untested."},{"line_number":41,"context_line":"Every instance save sends a notification of its update. That can, with the"},{"line_number":42,"context_line":"help of stacktach, actually be used to profile every large step of the boot"},{"line_number":43,"context_line":"process."},{"line_number":44,"context_line":""},{"line_number":45,"context_line":"Instance Faults"}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a0c5dc1_b8e8a5e3","line":42,"in_reply_to":"9a0c5dc1_4db9f2e8","updated":"2015-05-13 10:50:56.000000000","message":"yeah.","commit_id":"a09ffbf34e078e9ef1b755130d23594d1cc62ebe"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"d9f067c00040ea3f6864ff4109e3b26c4dc9b008","unresolved":false,"context_lines":[{"line_number":40,"context_line":"for billing. More worryingly they are unversioned and (largely) untested."},{"line_number":41,"context_line":"Every instance save sends a notification of its update. That can, with the"},{"line_number":42,"context_line":"help of stacktach, actually be used to profile every large step of the boot"},{"line_number":43,"context_line":"process."},{"line_number":44,"context_line":""},{"line_number":45,"context_line":"Instance Faults"},{"line_number":46,"context_line":"----------------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a0c5dc1_0d93fa5c","line":43,"updated":"2015-05-13 08:31:54.000000000","message":"One improvement point is how we can be sure that the notification formats are well defined. We have specific Exceptions, but notification payloads are just unversioned dicts AFAIK.","commit_id":"a09ffbf34e078e9ef1b755130d23594d1cc62ebe"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"779625aae2b1eb0d0392b6afc87e62ceca4236b2","unresolved":false,"context_lines":[{"line_number":40,"context_line":"for billing. More worryingly they are unversioned and (largely) untested."},{"line_number":41,"context_line":"Every instance save sends a notification of its update. That can, with the"},{"line_number":42,"context_line":"help of stacktach, actually be used to profile every large step of the boot"},{"line_number":43,"context_line":"process."},{"line_number":44,"context_line":""},{"line_number":45,"context_line":"Instance Faults"},{"line_number":46,"context_line":"----------------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a0c5dc1_38bb95e7","line":43,"in_reply_to":"9a0c5dc1_0d93fa5c","updated":"2015-05-13 10:50:56.000000000","message":"agreed, and actually testing them.","commit_id":"a09ffbf34e078e9ef1b755130d23594d1cc62ebe"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"d9f067c00040ea3f6864ff4109e3b26c4dc9b008","unresolved":false,"context_lines":[{"line_number":48,"context_line":"We store various trackbacks for errors that occur in the database."},{"line_number":49,"context_line":"There is no API to access these, but reading the database directly they can"},{"line_number":50,"context_line":"give you some good insight into issues that happen during the lifetime of"},{"line_number":51,"context_line":"each instance."},{"line_number":52,"context_line":""},{"line_number":53,"context_line":"Instance actions"},{"line_number":54,"context_line":"-----------------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a0c5dc1_6dddd641","line":51,"updated":"2015-05-13 08:31:54.000000000","message":"Agreed, a good way to mitigate those would be to access those in the API IMHO. Again, the problem is how you match your logs and your DB instance fault.","commit_id":"a09ffbf34e078e9ef1b755130d23594d1cc62ebe"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"779625aae2b1eb0d0392b6afc87e62ceca4236b2","unresolved":false,"context_lines":[{"line_number":48,"context_line":"We store various trackbacks for errors that occur in the database."},{"line_number":49,"context_line":"There is no API to access these, but reading the database directly they can"},{"line_number":50,"context_line":"give you some good insight into issues that happen during the lifetime of"},{"line_number":51,"context_line":"each instance."},{"line_number":52,"context_line":""},{"line_number":53,"context_line":"Instance actions"},{"line_number":54,"context_line":"-----------------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a0c5dc1_18be91d4","line":51,"in_reply_to":"9a0c5dc1_6dddd641","updated":"2015-05-13 10:50:56.000000000","message":"-1 but I need more context to explain why, coming soon.","commit_id":"a09ffbf34e078e9ef1b755130d23594d1cc62ebe"}]}