diff --git a/docs/speed/bot_health_sheriffing/how_to_access_test_logs.md b/docs/speed/bot_health_sheriffing/how_to_access_test_logs.md index 3fe2227bd32cec..5cb817c89f922e 100644 --- a/docs/speed/bot_health_sheriffing/how_to_access_test_logs.md +++ b/docs/speed/bot_health_sheriffing/how_to_access_test_logs.md @@ -10,29 +10,29 @@ When trying to understand a failure, it can be useful to inspect the test logs w Usually, Sheriff-o-matic will include a link directly in the alert to the most recent instance of the test failure. Just click on "shard #0 (failed)". -![Sheriff-o-matic shard #0 failed link](som_shard_0.png) +![Sheriff-o-matic shard #0 failed link](images/som_shard_0.png) ### Accessing the logs for older failures From the Sheriff-o-matic alert, click the "Examine" link to access a list of recent runs on the given bot. -![Sheriff-o-matic open examine pane](som_examine.png) +![Sheriff-o-matic open examine pane](images/som_examine.png) If the failure spans multiple platforms, you can select which platform you care about from the top of the new pane. -![Sheriff-o-matic choose bot from examine pane](som_examine_choose_bot.png) +![Sheriff-o-matic choose bot from examine pane](images/som_examine_choose_bot.png) This new pane shows all recent runs, not just runs where this benchmark failed, so it's often useful to use Ctrl+F and search for the failing benchmark name to highlight these runs.Next, click the build number for the run for which you want to access the logs. Once you've found the run you want the logs for, click the build number for the run for which you want to access the logs. -![Sheriff-o-matic choose build number](som_choose_build_number.png) +![Sheriff-o-matic choose build number](images/som_choose_build_number.png) Once at the build page listing all steps, we need to find the test step that failed. Start by selecting to show "Non-Green Only" steps at the top of the page. -![Sheriff-o-matic choose non-green only](som_choose_non_green_only.png) +![Sheriff-o-matic choose non-green only](images/som_choose_non_green_only.png) After doing this, search for your benchmark's name (in this case, "v8.browsing_desktop") until you find a step with a name like " on ". Click the "shard #0 (failed)" link below that step to open the logs. -![Sheriff-o-matic choose shard #0 failed link from test steps](som_test_steps_shard_0.png) +![Sheriff-o-matic choose shard #0 failed link from test steps](images/som_test_steps_shard_0.png) ## Navigating log files diff --git a/docs/speed/bot_health_sheriffing/how_to_address_duplicate_alerts.md b/docs/speed/bot_health_sheriffing/how_to_address_duplicate_alerts.md index 758f5f6c1e8f1e..7ee499080301e7 100644 --- a/docs/speed/bot_health_sheriffing/how_to_address_duplicate_alerts.md +++ b/docs/speed/bot_health_sheriffing/how_to_address_duplicate_alerts.md @@ -10,17 +10,17 @@ First, decide which of your duplicate alerts is going to be your "main" alert. I Next, select all of the other alerts that you want to merge into this main alert. If there are any existing alert groups among these other alerts, you need to click "Ungroup all" and check all of the boxes in the dialog to ensure that the alerts are out of their existing groups. -![Sheriff-o-matic ungroup all](ungroup_all.png) +![Sheriff-o-matic ungroup all](images/ungroup_all.png) -![Sheriff-o-matic bulk ungroup dialog](bulk_ungroup.png) +![Sheriff-o-matic bulk ungroup dialog](images/bulk_ungroup.png) Once you've ungrouped all the alerts being merged into the main alert and selected all alerts being merged, including your primary alert, click the "Group all" button at the top of the screen. -![Sheriff-o-matic group all](group_all.png) +![Sheriff-o-matic group all](images/group_all.png) If the alert group still has an auto-generated name, it's a good idea to give it a name that clarifies the problem. -![Changing the group name in Sheriff-o-matic](change_group_name.png) +![Changing the group name in Sheriff-o-matic](images/change_group_name.png) Lastly, it may be necessary to broaden the scope of the existing bug. If, for example, the previous alert had the name "system_health.common_desktop failing on Mac Retina Perf" and you merged in a duplicate alert on Mac Air Perf, you should change the bug and alert names to "system_health.common_desktop failing on multiple Mac builders" and simultaneously add a comment to the bug to the effect of "I'm broadening this bug because the issue is also causing failures on Mac Air Perf". diff --git a/docs/speed/bot_health_sheriffing/how_to_snooze_an_alert.md b/docs/speed/bot_health_sheriffing/how_to_snooze_an_alert.md index f19f58ca3641d5..458d337e95fd5d 100644 --- a/docs/speed/bot_health_sheriffing/how_to_snooze_an_alert.md +++ b/docs/speed/bot_health_sheriffing/how_to_snooze_an_alert.md @@ -4,8 +4,8 @@ After addressing an alert, the next step is to snooze it. Snoozing an alert hi To snooze an alert, click the "Snooze" button for the specified alert and then enter the number of minutes you'd like to snooze the alert for. -![Snooze an alert](snooze_alert.png) +![Snooze an alert](images/snooze_alert.png) -![Alert snooze dialog](snooze_alert_dialog.png) +![Alert snooze dialog](images/snooze_alert_dialog.png) Generally, you should snooze an alert for however long you think it will take for your fix to take effect and a successful run to complete, removing the alert from Sheriff-o-matic. If you're unsure of how long this should be, 24 hours (1440 minutes) is usually a safe default. diff --git a/docs/speed/bot_health_sheriffing/bulk_ungroup.png b/docs/speed/bot_health_sheriffing/images/bulk_ungroup.png similarity index 100% rename from docs/speed/bot_health_sheriffing/bulk_ungroup.png rename to docs/speed/bot_health_sheriffing/images/bulk_ungroup.png diff --git a/docs/speed/bot_health_sheriffing/change_group_name.png b/docs/speed/bot_health_sheriffing/images/change_group_name.png similarity index 100% rename from docs/speed/bot_health_sheriffing/change_group_name.png rename to docs/speed/bot_health_sheriffing/images/change_group_name.png diff --git a/docs/speed/bot_health_sheriffing/flakiness_dashboard.png b/docs/speed/bot_health_sheriffing/images/flakiness_dashboard.png similarity index 100% rename from docs/speed/bot_health_sheriffing/flakiness_dashboard.png rename to docs/speed/bot_health_sheriffing/images/flakiness_dashboard.png diff --git a/docs/speed/bot_health_sheriffing/group_all.png b/docs/speed/bot_health_sheriffing/images/group_all.png similarity index 100% rename from docs/speed/bot_health_sheriffing/group_all.png rename to docs/speed/bot_health_sheriffing/images/group_all.png diff --git a/docs/speed/bot_health_sheriffing/snooze_alert.png b/docs/speed/bot_health_sheriffing/images/snooze_alert.png similarity index 100% rename from docs/speed/bot_health_sheriffing/snooze_alert.png rename to docs/speed/bot_health_sheriffing/images/snooze_alert.png diff --git a/docs/speed/bot_health_sheriffing/snooze_alert_dialog.png b/docs/speed/bot_health_sheriffing/images/snooze_alert_dialog.png similarity index 100% rename from docs/speed/bot_health_sheriffing/snooze_alert_dialog.png rename to docs/speed/bot_health_sheriffing/images/snooze_alert_dialog.png diff --git a/docs/speed/bot_health_sheriffing/som_choose_build_number.png b/docs/speed/bot_health_sheriffing/images/som_choose_build_number.png similarity index 100% rename from docs/speed/bot_health_sheriffing/som_choose_build_number.png rename to docs/speed/bot_health_sheriffing/images/som_choose_build_number.png diff --git a/docs/speed/bot_health_sheriffing/som_choose_non_green_only.png b/docs/speed/bot_health_sheriffing/images/som_choose_non_green_only.png similarity index 100% rename from docs/speed/bot_health_sheriffing/som_choose_non_green_only.png rename to docs/speed/bot_health_sheriffing/images/som_choose_non_green_only.png diff --git a/docs/speed/bot_health_sheriffing/som_examine.png b/docs/speed/bot_health_sheriffing/images/som_examine.png similarity index 100% rename from docs/speed/bot_health_sheriffing/som_examine.png rename to docs/speed/bot_health_sheriffing/images/som_examine.png diff --git a/docs/speed/bot_health_sheriffing/som_examine_choose_bot.png b/docs/speed/bot_health_sheriffing/images/som_examine_choose_bot.png similarity index 100% rename from docs/speed/bot_health_sheriffing/som_examine_choose_bot.png rename to docs/speed/bot_health_sheriffing/images/som_examine_choose_bot.png diff --git a/docs/speed/bot_health_sheriffing/som_shard_0.png b/docs/speed/bot_health_sheriffing/images/som_shard_0.png similarity index 100% rename from docs/speed/bot_health_sheriffing/som_shard_0.png rename to docs/speed/bot_health_sheriffing/images/som_shard_0.png diff --git a/docs/speed/bot_health_sheriffing/som_test_steps_shard_0.png b/docs/speed/bot_health_sheriffing/images/som_test_steps_shard_0.png similarity index 100% rename from docs/speed/bot_health_sheriffing/som_test_steps_shard_0.png rename to docs/speed/bot_health_sheriffing/images/som_test_steps_shard_0.png diff --git a/docs/speed/bot_health_sheriffing/ungroup_all.png b/docs/speed/bot_health_sheriffing/images/ungroup_all.png similarity index 100% rename from docs/speed/bot_health_sheriffing/ungroup_all.png rename to docs/speed/bot_health_sheriffing/images/ungroup_all.png diff --git a/docs/speed/bot_health_sheriffing/what_test_is_failing.md b/docs/speed/bot_health_sheriffing/what_test_is_failing.md index 34317d40b2bd72..9eb5b192722beb 100644 --- a/docs/speed/bot_health_sheriffing/what_test_is_failing.md +++ b/docs/speed/bot_health_sheriffing/what_test_is_failing.md @@ -4,7 +4,7 @@ The first step in addressing a test failure is to identify what stories are fail The easiest way to identify these is to use the [Flakiness dashboard](https://test-results.appspot.com/dashboards/flakiness_dashboard.html#testType=blink_perf.canvas), which is a high-level dashboard showing test passes and failures. (Sheriff-o-matic tries to automatically identify the failing stories, but is often incorrect and therefore can't be trusted.) Open up the flakiness dashboard and select the benchmark and platform in question (pulled from the SOM alert) from the "Test type" and "Builder" dropdowns. You should see a view like this: -![The flakiness dashboard](flakiness_dashboard.png) +![The flakiness dashboard](images/flakiness_dashboard.png) Each row represents a particular story and each column represents a recent run, listed with the most recent run on the left. If the cell is green, then the story passed; if it's red, then it failed. Only stories that have failed at least once will be listed. You can click on a particular cell to see more information like revision ranges (useful for launching bisects) and logs.