Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

fix: improve the time filters in DORA dashboards #8114

Merged
merged 1 commit into from
Sep 29, 2024
Merged

fix: improve the time filters in DORA dashboards #8114

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions grafana/dashboards/DORA.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -962,7 +962,7 @@
"metricColumn": "none",
"queryType": "randomWalk",
"rawQuery": true,
"rawSql": "-- Metric 1: Number of deployments per month\nwith _deployments as(\n-- When deploying multiple commits in one pipeline, GitLab and BitBucket may generate more than one deployment. However, DevLake consider these deployments as ONE production deployment and use the last one's finished_date as the finished date.\n\tSELECT \n\t\tdate_format(deployment_finished_date,'%y/%m') as month,\n\t\tcount(cicd_deployment_id) as deployment_count\n\tFROM (\n\t\tSELECT\n\t\t\tcdc.cicd_deployment_id,\n\t\t\tmax(cdc.finished_date) as deployment_finished_date\n\t\tFROM cicd_deployment_commits cdc\n\t\tJOIN project_mapping pm on cdc.cicd_scope_id = pm.row_id and pm.`table` = 'cicd_scopes'\n\t\tWHERE\n\t\t\tpm.project_name in (${project})\n\t\t\tand cdc.result = 'SUCCESS'\n\t\t\tand cdc.environment = 'PRODUCTION'\n\t\tGROUP BY 1\n\t\tHAVING $__timeFilter(max(cdc.finished_date))\n\t) _production_deployments\n\tGROUP BY 1\n)\n\nSELECT \n\tcm.month, \n\tcase when d.deployment_count is null then 0 else d.deployment_count end as 'Deployment Count'\nFROM \n\tcalendar_months cm\n\tLEFT JOIN _deployments d on cm.month = d.month\n\tWHERE $__timeFilter(cm.month_timestamp)",
"rawSql": "-- Metric 1: Number of deployments per month\nwith _deployments as(\n-- When deploying multiple commits in one pipeline, GitLab and BitBucket may generate more than one deployment. However, DevLake consider these deployments as ONE production deployment and use the last one's finished_date as the finished date.\n\tSELECT \n\t\tdate_format(deployment_finished_date,'%y/%m') as month,\n\t\tcount(cicd_deployment_id) as deployment_count\n\tFROM (\n\t\tSELECT\n\t\t\tcdc.cicd_deployment_id,\n\t\t\tmax(cdc.finished_date) as deployment_finished_date\n\t\tFROM cicd_deployment_commits cdc\n\t\tJOIN project_mapping pm on cdc.cicd_scope_id = pm.row_id and pm.`table` = 'cicd_scopes'\n\t\tWHERE\n\t\t\tpm.project_name in (${project})\n\t\t\tand cdc.result = 'SUCCESS'\n\t\t\tand cdc.environment = 'PRODUCTION'\n\t\tGROUP BY 1\n\t\tHAVING $__timeFilter(max(cdc.finished_date))\n\t) _production_deployments\n\tGROUP BY 1\n)\n\nSELECT \n\tcm.month, \n\tcase when d.deployment_count is null then 0 else d.deployment_count end as 'Deployment Count'\nFROM \n\tcalendar_months cm\n\tLEFT JOIN _deployments d on cm.month = d.month\n\tWHERE month_timestamp between DATE(DATE_FORMAT($__timeFrom(), '%Y-%m-01')) AND DATE(DATE_FORMAT($__timeTo(), '%Y-%m-01'))",
"refId": "A",
"select": [
[
Expand Down Expand Up @@ -1095,7 +1095,7 @@
"hide": false,
"metricColumn": "none",
"rawQuery": true,
"rawSql": "-- Metric 2: median change lead time per month\nwith _pr_stats as (\n-- get the cycle time of PRs deployed by the deployments finished each month\n\tSELECT\n\t\tdistinct pr.id,\n\t\tdate_format(cdc.finished_date,'%y/%m') as month,\n\t\tppm.pr_cycle_time\n\tFROM\n\t\tpull_requests pr\n\t\tjoin project_pr_metrics ppm on ppm.id = pr.id\n\t\tjoin project_mapping pm on pr.base_repo_id = pm.row_id and pm.`table` = 'repos'\n\t\tjoin cicd_deployment_commits cdc on ppm.deployment_commit_id = cdc.id\n\tWHERE\n\t\tpm.project_name in (${project}) \n\t\tand pr.merged_date is not null\n\t\tand ppm.pr_cycle_time is not null\n\t\tand $__timeFilter(cdc.finished_date)\n),\n\n_find_median_clt_each_month_ranks as(\n\tSELECT *, percent_rank() over(PARTITION BY month order by pr_cycle_time) as ranks\n\tFROM _pr_stats\n),\n\n_clt as(\n\tSELECT month, max(pr_cycle_time) as median_change_lead_time\n\tFROM _find_median_clt_each_month_ranks\n\tWHERE ranks <= 0.5\n\tgroup by month\n)\n\nSELECT \n\tcm.month,\n\tcase \n\t\twhen _clt.median_change_lead_time is null then 0 \n\t\telse _clt.median_change_lead_time/60 end as 'Median Change Lead Time In Hours'\nFROM \n\tcalendar_months cm\n\tLEFT JOIN _clt on cm.month = _clt.month\n WHERE $__timeFilter(cm.month_timestamp)",
"rawSql": "-- Metric 2: median change lead time per month\nwith _pr_stats as (\n-- get the cycle time of PRs deployed by the deployments finished each month\n\tSELECT\n\t\tdistinct pr.id,\n\t\tdate_format(cdc.finished_date,'%y/%m') as month,\n\t\tppm.pr_cycle_time\n\tFROM\n\t\tpull_requests pr\n\t\tjoin project_pr_metrics ppm on ppm.id = pr.id\n\t\tjoin project_mapping pm on pr.base_repo_id = pm.row_id and pm.`table` = 'repos'\n\t\tjoin cicd_deployment_commits cdc on ppm.deployment_commit_id = cdc.id\n\tWHERE\n\t\tpm.project_name in (${project}) \n\t\tand pr.merged_date is not null\n\t\tand ppm.pr_cycle_time is not null\n\t\tand $__timeFilter(cdc.finished_date)\n),\n\n_find_median_clt_each_month_ranks as(\n\tSELECT *, percent_rank() over(PARTITION BY month order by pr_cycle_time) as ranks\n\tFROM _pr_stats\n),\n\n_clt as(\n\tSELECT month, max(pr_cycle_time) as median_change_lead_time\n\tFROM _find_median_clt_each_month_ranks\n\tWHERE ranks <= 0.5\n\tgroup by month\n)\n\nSELECT \n\tcm.month,\n\tcase \n\t\twhen _clt.median_change_lead_time is null then 0 \n\t\telse _clt.median_change_lead_time/60 end as 'Median Change Lead Time In Hours'\nFROM \n\tcalendar_months cm\n\tLEFT JOIN _clt on cm.month = _clt.month\n WHERE month_timestamp between DATE(DATE_FORMAT($__timeFrom(), '%Y-%m-01')) AND DATE(DATE_FORMAT($__timeTo(), '%Y-%m-01'))",
"refId": "A",
"select": [
[
Expand Down Expand Up @@ -1249,7 +1249,7 @@
"hide": false,
"metricColumn": "none",
"rawQuery": true,
"rawSql": "-- Metric 3: change failure rate per month\nwith _deployments as (\n -- When deploying multiple commits in one pipeline, GitLab and BitBucket may generate more than one deployment. However, DevLake consider these deployments as ONE production deployment and use the last one's finished_date as the finished date.\n SELECT\n cdc.cicd_deployment_id as deployment_id,\n max(cdc.finished_date) as deployment_finished_date\n FROM\n cicd_deployment_commits cdc\n JOIN project_mapping pm on cdc.cicd_scope_id = pm.row_id\n and pm.`table` = 'cicd_scopes'\n WHERE\n pm.project_name in (${project})\n and cdc.result = 'SUCCESS'\n and cdc.environment = 'PRODUCTION'\n GROUP BY\n 1\n HAVING\n $__timeFilter(max(cdc.finished_date))\n),\n_failure_caused_by_deployments as (\n -- calculate the number of incidents caused by each deployment\n SELECT\n d.deployment_id,\n d.deployment_finished_date,\n count(\n distinct case\n when i.id is not null then d.deployment_id\n else null\n end\n ) as has_incident\n FROM\n _deployments d\n left join project_incident_deployment_relationships pim on d.deployment_id = pim.deployment_id\n left join incidents i on pim.id = i.id\n GROUP BY\n 1,\n 2\n),\n_change_failure_rate_for_each_month as (\n SELECT\n date_format(deployment_finished_date, '%y/%m') as month,\n case\n when count(deployment_id) is null then null\n else sum(has_incident) / count(deployment_id)\n end as change_failure_rate\n FROM\n _failure_caused_by_deployments\n GROUP BY\n 1\n)\nSELECT\n cm.month,\n cfr.change_failure_rate as 'Change Failure Rate'\nFROM\n calendar_months cm\n LEFT JOIN _change_failure_rate_for_each_month cfr on cm.month = cfr.month\nWHERE\n $__timeFilter(cm.month_timestamp)",
"rawSql": "-- Metric 3: change failure rate per month\nwith _deployments as (\n -- When deploying multiple commits in one pipeline, GitLab and BitBucket may generate more than one deployment. However, DevLake consider these deployments as ONE production deployment and use the last one's finished_date as the finished date.\n SELECT\n cdc.cicd_deployment_id as deployment_id,\n max(cdc.finished_date) as deployment_finished_date\n FROM\n cicd_deployment_commits cdc\n JOIN project_mapping pm on cdc.cicd_scope_id = pm.row_id\n and pm.`table` = 'cicd_scopes'\n WHERE\n pm.project_name in (${project})\n and cdc.result = 'SUCCESS'\n and cdc.environment = 'PRODUCTION'\n GROUP BY\n 1\n HAVING\n $__timeFilter(max(cdc.finished_date))\n),\n_failure_caused_by_deployments as (\n -- calculate the number of incidents caused by each deployment\n SELECT\n d.deployment_id,\n d.deployment_finished_date,\n count(\n distinct case\n when i.id is not null then d.deployment_id\n else null\n end\n ) as has_incident\n FROM\n _deployments d\n left join project_incident_deployment_relationships pim on d.deployment_id = pim.deployment_id\n left join incidents i on pim.id = i.id\n GROUP BY\n 1,\n 2\n),\n_change_failure_rate_for_each_month as (\n SELECT\n date_format(deployment_finished_date, '%y/%m') as month,\n case\n when count(deployment_id) is null then null\n else sum(has_incident) / count(deployment_id)\n end as change_failure_rate\n FROM\n _failure_caused_by_deployments\n GROUP BY\n 1\n)\nSELECT\n cm.month,\n cfr.change_failure_rate as 'Change Failure Rate'\nFROM\n calendar_months cm\n LEFT JOIN _change_failure_rate_for_each_month cfr on cm.month = cfr.month\nWHERE\n month_timestamp between DATE(DATE_FORMAT($__timeFrom(), '%Y-%m-01')) AND DATE(DATE_FORMAT($__timeTo(), '%Y-%m-01'))",
"refId": "A",
"select": [
[
Expand Down Expand Up @@ -1407,7 +1407,7 @@
"hide": false,
"metricColumn": "none",
"rawQuery": true,
"rawSql": "-- ***** 2023 report ***** --\n-- Metric 4: Failed deployment recovery time\nwith _deployments as (\n SELECT\n cdc.cicd_deployment_id as deployment_id,\n max(cdc.finished_date) as deployment_finished_date\n FROM\n cicd_deployment_commits cdc\n JOIN project_mapping pm on cdc.cicd_scope_id = pm.row_id\n and pm.`table` = 'cicd_scopes'\n WHERE\n pm.project_name in ($project)\n and cdc.result = 'SUCCESS'\n and cdc.environment = 'PRODUCTION'\n GROUP BY\n 1\n HAVING\n $__timeFilter(max(cdc.finished_date))\n),\n_incidents_for_deployments as (\n SELECT\n i.id as incident_id,\n i.created_date as incident_create_date,\n i.resolution_date as incident_resolution_date,\n fd.deployment_id as caused_by_deployment,\n fd.deployment_finished_date,\n date_format(fd.deployment_finished_date, '%y/%m') as deployment_finished_month\n FROM\n incidents i\n left join project_incident_deployment_relationships pim on i.id = pim.id\n join _deployments fd on pim.deployment_id = fd.deployment_id\n WHERE\n $__timeFilter(i.resolution_date)\n),\n_recovery_time_ranks as (\n SELECT\n *,\n percent_rank() over(\n PARTITION BY deployment_finished_month\n order by\n TIMESTAMPDIFF(\n MINUTE,\n deployment_finished_date,\n incident_resolution_date\n )\n ) as ranks\n FROM\n _incidents_for_deployments\n),\n_median_recovery_time as (\n SELECT\n deployment_finished_month,\n max(\n TIMESTAMPDIFF(\n MINUTE,\n deployment_finished_date,\n incident_resolution_date\n )\n ) as median_recovery_time\n FROM\n _recovery_time_ranks\n WHERE\n ranks <= 0.5\n GROUP BY\n deployment_finished_month\n),\n_metric_recovery_time_2023_report as (\n SELECT\n cm.month,\n case\n when m.median_recovery_time is null then 0\n else m.median_recovery_time / 60\n end as median_recovery_time_in_hour\n FROM\n calendar_months cm\n LEFT JOIN _median_recovery_time m on cm.month = m.deployment_finished_month\n WHERE\n $__timeFilter(cm.month_timestamp)\n),\n-- ***** 2021 report ***** --\n-- Metric 4: median time to restore service - MTTR\n_incidents as (\n -- get the number of incidents created each month\n SELECT\n distinct i.id,\n date_format(i.resolution_date, '%y/%m') as month,\n cast(lead_time_minutes as signed) as lead_time_minutes\n FROM\n incidents i\n join project_mapping pm on i.scope_id = pm.row_id\n and pm.`table` = i.`table`\n WHERE\n pm.project_name in (${project})\n and i.lead_time_minutes is not null\n),\n_find_median_mttr_each_month_ranks as(\n SELECT\n *,\n percent_rank() over(\n PARTITION BY month\n order by\n lead_time_minutes\n ) as ranks\n FROM\n _incidents\n),\n_mttr as(\n SELECT\n month,\n max(lead_time_minutes) as median_time_to_resolve\n FROM\n _find_median_mttr_each_month_ranks\n WHERE\n ranks <= 0.5\n GROUP BY\n month\n),\n_metric_mttr_2021_report as (\n SELECT\n cm.month,\n case\n when m.median_time_to_resolve is null then 0\n else m.median_time_to_resolve / 60\n end as median_time_to_resolve_in_hour\n FROM\n calendar_months cm\n LEFT JOIN _mttr m on cm.month = m.month\n WHERE\n $__timeFilter(cm.month_timestamp)\n)\nSELECT\n cm.month,\n CASE\n WHEN '${dora_report}' = '2023' THEN mrt.median_recovery_time_in_hour\n WHEN '${dora_report}' = '2021' THEN mm.median_time_to_resolve_in_hour\n END AS '${title_value} In Hours'\nFROM\n calendar_months cm\n LEFT JOIN _metric_recovery_time_2023_report mrt ON cm.month = mrt.month\n LEFT JOIN _metric_mttr_2021_report mm ON cm.month = mm.month\nWHERE\n $__timeFilter(cm.month_timestamp)",
"rawSql": "-- ***** 2023 report ***** --\n-- Metric 4: Failed deployment recovery time\nwith _deployments as (\n SELECT\n cdc.cicd_deployment_id as deployment_id,\n max(cdc.finished_date) as deployment_finished_date\n FROM\n cicd_deployment_commits cdc\n JOIN project_mapping pm on cdc.cicd_scope_id = pm.row_id\n and pm.`table` = 'cicd_scopes'\n WHERE\n pm.project_name in ($project)\n and cdc.result = 'SUCCESS'\n and cdc.environment = 'PRODUCTION'\n GROUP BY\n 1\n HAVING\n $__timeFilter(max(cdc.finished_date))\n),\n_incidents_for_deployments as (\n SELECT\n i.id as incident_id,\n i.created_date as incident_create_date,\n i.resolution_date as incident_resolution_date,\n fd.deployment_id as caused_by_deployment,\n fd.deployment_finished_date,\n date_format(fd.deployment_finished_date, '%y/%m') as deployment_finished_month\n FROM\n incidents i\n left join project_incident_deployment_relationships pim on i.id = pim.id\n join _deployments fd on pim.deployment_id = fd.deployment_id\n WHERE\n $__timeFilter(i.resolution_date)\n),\n_recovery_time_ranks as (\n SELECT\n *,\n percent_rank() over(\n PARTITION BY deployment_finished_month\n order by\n TIMESTAMPDIFF(\n MINUTE,\n deployment_finished_date,\n incident_resolution_date\n )\n ) as ranks\n FROM\n _incidents_for_deployments\n),\n_median_recovery_time as (\n SELECT\n deployment_finished_month,\n max(\n TIMESTAMPDIFF(\n MINUTE,\n deployment_finished_date,\n incident_resolution_date\n )\n ) as median_recovery_time\n FROM\n _recovery_time_ranks\n WHERE\n ranks <= 0.5\n GROUP BY\n deployment_finished_month\n),\n_metric_recovery_time_2023_report as (\n SELECT\n cm.month,\n case\n when m.median_recovery_time is null then 0\n else m.median_recovery_time / 60\n end as median_recovery_time_in_hour\n FROM\n calendar_months cm\n LEFT JOIN _median_recovery_time m on cm.month = m.deployment_finished_month\n WHERE\n month_timestamp between DATE(DATE_FORMAT($__timeFrom(), '%Y-%m-01')) AND DATE(DATE_FORMAT($__timeTo(), '%Y-%m-01'))\n),\n-- ***** 2021 report ***** --\n-- Metric 4: median time to restore service - MTTR\n_incidents as (\n -- get the number of incidents created each month\n SELECT\n distinct i.id,\n date_format(i.resolution_date, '%y/%m') as month,\n cast(lead_time_minutes as signed) as lead_time_minutes\n FROM\n incidents i\n join project_mapping pm on i.scope_id = pm.row_id\n and pm.`table` = i.`table`\n WHERE\n pm.project_name in (${project})\n and i.lead_time_minutes is not null\n),\n_find_median_mttr_each_month_ranks as(\n SELECT\n *,\n percent_rank() over(\n PARTITION BY month\n order by\n lead_time_minutes\n ) as ranks\n FROM\n _incidents\n),\n_mttr as(\n SELECT\n month,\n max(lead_time_minutes) as median_time_to_resolve\n FROM\n _find_median_mttr_each_month_ranks\n WHERE\n ranks <= 0.5\n GROUP BY\n month\n),\n_metric_mttr_2021_report as (\n SELECT\n cm.month,\n case\n when m.median_time_to_resolve is null then 0\n else m.median_time_to_resolve / 60\n end as median_time_to_resolve_in_hour\n FROM\n calendar_months cm\n LEFT JOIN _mttr m on cm.month = m.month\n WHERE\n month_timestamp between DATE(DATE_FORMAT($__timeFrom(), '%Y-%m-01')) AND DATE(DATE_FORMAT($__timeTo(), '%Y-%m-01'))\n)\nSELECT\n cm.month,\n CASE\n WHEN '${dora_report}' = '2023' THEN mrt.median_recovery_time_in_hour\n WHEN '${dora_report}' = '2021' THEN mm.median_time_to_resolve_in_hour\n END AS '${title_value} In Hours'\nFROM\n calendar_months cm\n LEFT JOIN _metric_recovery_time_2023_report mrt ON cm.month = mrt.month\n LEFT JOIN _metric_mttr_2021_report mm ON cm.month = mm.month\nWHERE\n month_timestamp between DATE(DATE_FORMAT($__timeFrom(), '%Y-%m-01')) AND DATE(DATE_FORMAT($__timeTo(), '%Y-%m-01'))",
"refId": "A",
"select": [
[
Expand Down
Loading
Loading