3/2022; feature launched. manukh@chromium.org jdonnelly@chromium.org chrome-omnibox-team@google.com The time elapsed decoding an image using BitmapFetcher. 3/2022; feature launched. manukh@chromium.org jdonnelly@chromium.org chrome-omnibox-team@google.com The time elapsed fetching an image from a URL using BitmapFetcher. M109; Replaced by "ChromeOS.HaTS.*". aalsum@chromium.org jackshira@google.org cros-telemetry@google.com Records the overall satisfaction score for HaTS, to be analyzed against other OS metrics like performance or number of crashes. There is one score per survey type. Emitted when a user answers the first question on a HaTS survey. aalsum@chromium.org cros-telemetry@google.com Records the stage up to which the user interacted with the HaTS Chrome OS survey. This is needed in addition to the Hats metrics data since Chrome OS relies on client side logic to roll the dice for device selection. This is recorded when the Hats pipeline ends processing, which is usually when the controller is destroyed. lgrey@chromium.org robliao@chromium.org Whether the user's system is in dark mode, light mode, or if this choice is unavailable due to lack of system support. Updated every UMA upload. vshenvi@google.com cros-reporting-team@google.com Recorded when the AppEventsObserver::AppInstallTracker starts tracking new app installs or uninstalls. This represents the total disk space consumed by the component in the user pref store before encryption. albertojuarez@google.com src/components/reporting/OWNERS Recorded when RecordHandlerImpl catches an error while trying to parse the configuration file that was sent from the reporting server. lbaraz@chromium.org xuhong@chromium.org src/components/reporting/OWNERS Recorded when ReportQueue::Enqueue is called to post an event to ERP. It counts occurrences of the call returning success or each possible error Status. xuhong@chromium.org lbaraz@chromium.org cros-reporting-team@google.com {Param} informed by the Missive daemon. Updated every time the Missive daemon requests uploading. lbaraz@chromium.org cros-reporting-team@google.com Records whether an event is filtered out by FilteredReportQueue. xuhong@chromium.org cros-reporting-team@google.com Records the result of obtaining PSR info from healthd in the metrics pipeline. lbaraz@chromium.org cros-reporting-team@google.com Records whether an event is accepted by ReportQueue rate limiter. jrhilke@chromium.org lbaraz@chromium.org cros-reporting-team@google.com Recorded when uploading records to the reporting server. Counts the number of records batched into a single upload request from a {ManagementStatus} device. Uploads containing zero records will not be included as those are just encryption key requests. xuhong@chromium.org lbaraz@chromium.org cros-reporting-team@google.com The size of ERP's response payload as a JSON string when upload is successful. Updated every time when an upload succeeds AND at least one hour has passed since the last update AND 9 more uploads have succeeded since the last update. xuhong@chromium.org lbaraz@chromium.org cros-reporting-team@google.com The size of ERP's request payload as a JSON string when upload is successful. Updated every time when an upload succeeds. vshenvi@google.com cros-reporting-team@google.com Would only be true. Recorded whenever the generation id is reset in the storage queue. The generation id is set up once during storage queue initialization using the metadata file that matches the last sequencing id used to track, deduplicate and confirm delivery by the server. The generation id is reset when this file is missing or corrupted and the sequencing is irreparably broken. xuhong@chromium.org lbaraz@chromium.org cros-reporting-team@google.com Would only be true. Recorded whenever an unusual timestamp is encountered when enqueuing a record, which will be rejected. jrhilke@google.com src/components/reporting/OWNERS Status of a completed ERP upload request from a {ManagementStatus}. Either successful, with a failure or with a server error. Recorded when a response from the server is received. xuhong@chromium.org lbaraz@chromium.org cros-reporting-team@google.com The accumulated size per hour of ERP's {Param} payload as JSON strings when upload is successful. Updated every hour. pmonette@chromium.org catan-team@chromium.org This metric is emitted every 30 seconds after main message loop start, when there is user activity. Each 30 second duration is divided into 100ms intervals. This metric counts the number of these intervals that were "congested". An interval is congested if during it the UI or IO thread executes an "important" task that was queued more than 100 ms ago, or a non-important task or a native event handler that started executing more than 100 ms ago. See https://docs.google.com/document/d/1vDSGFvJblh7yJ3U3RVB_7qZLubyfTbQdQjuN1GoUNkc/edit for more details. This metric is only recorded if Chrome is either visible, playing audio, or capturing video. etienneb@chromium.org gab@chromium.org chrome-analysis-team@google.com This metric is emitted every 30 seconds after main message loop start, when there is user activity. Each 30 second duration is divided into 100ms intervals. This metric counts the number of these intervals that were "congested". An interval is congested if during it the UI or IO thread executes an "important" task that was queued more than 100 ms ago, or a non-important task or a native event handler that started executing more than 100 ms ago. See https://docs.google.com/document/d/1vDSGFvJblh7yJ3U3RVB_7qZLubyfTbQdQjuN1GoUNkc/edit for more details. This metrics graduated from Browser.Responsiveness.JankyIntervalsPerThirtySeconds(1,2,3) and their history might be useful to understand how we got here. Suffixed phases can be useful as diagnosis metrics. This phase represents: {Phase} This histogram is of special interest to the chrome-analysis-team@. Do not change its semantics or retire it without talking to them first. pmonette@chromium.org catan-team@chromium.org This metric is emitted every 30 seconds after main message loop start, when there is user activity. Each 30 second duration is divided into 100ms intervals. This metric counts the number of these intervals that were "congested". An interval is congested if during it the UI or IO thread executes an "important" task that was queued more than 100 ms ago, or a non-important task or a native event handler that started executing more than 100 ms ago. See https://docs.google.com/document/d/1vDSGFvJblh7yJ3U3RVB_7qZLubyfTbQdQjuN1GoUNkc/edit for more details. This metric is split by usage scenarios (go/chrome_power_use_per_scenario). This version is recorded for {UsageScenario}. ckitagawa@chromium.org fredmello@chromium.org chrome-fdt@google.com Records the size of a compressed on-disk capture for the paint preview after a proto is written to disk if compressed. ckitagawa@chromium.org fredmello@chromium.org chrome-fdt@google.com Records the number of frames captured by a paint preview if it succeeded. ckitagawa@chromium.org fredmello@chromium.org chrome-fdt@google.com Records a boolean indicating whether a capture attempt was successful. ckitagawa@chromium.org fredmello@chromium.org chrome-fdt@google.com Records the total time spent capturing a paint preview of a page. This includes capture and serialization of all frames and IPC time. ckitagawa@chromium.org fredmello@chromium.org chrome-fdt@google.com Records if the paint preview compositor process started correctly. ckitagawa@chromium.org fredmello@chromium.org chrome-fdt@google.com Records the time it took to start the paint preview compositor process. ckitagawa@chromium.org fredmello@chromium.org chrome-fdt@google.com Records when a link is clicked in the paint preview player. ckitagawa@chromium.org fredmello@chromium.org chrome-fdt@google.com Records the time taken to from the request to start the compositor process to the first bitmap being loaded. ckitagawa@chromium.org fredmello@chromium.org chrome-fdt@google.com Records the the reason for exiting the compositor process for the TabbedPaintPreviewPlayer. Recorded when the compositor returns an error or disconnects. ckitagawa@chromium.org fredmello@chromium.org chrome-fdt@google.com Records the the reason for exiting TabbedPaintPreviewPlayer. Recorded when paint preview is removed. ckitagawa@chromium.org fredmello@chromium.org chrome-fdt@google.com Records whether paint preview was drawn before the tab finished loading. Recorded when tab has finished loading. ckitagawa@chromium.org fredmello@chromium.org chrome-fdt@google.com Records whether a paint preview exists. Recorded when attempting to show a preview on startup. This in conjuction with CompositorFailureReason gives the total number of failures to show a paint preview. ckitagawa@chromium.org fredmello@chromium.org chrome-fdt@google.com Records the number of times that the upgrade snackbar was shown in a session of TabbedPaintPreviewPlayer. Recorded when paint preview is removed. ckitagawa@chromium.org fredmello@chromium.org chrome-fdt@google.com Records the time taken to from actvity creation to the first bitmap being shown. Recorded on first paint preview paint event. ckitagawa@chromium.org fredmello@chromium.org chrome-fdt@google.com Records the length of the time that a session of TabbedPaintPreviewPlayer lasts. Recorded when paint preview is removed. {TabbedPaintPreviewPlayerUptime} ckitagawa@chromium.org fredmello@chromium.org Records the total disk usage by the Paint Preview Tab Service at startup of the profile keyed service. gab@chromium.org olivierli@chromium.org Similar to Browser.Responsiveness.IOJankyIntervalsPerMinute but sums janky events across all foreground threads over the monitored minute rather than considering each interval a boolean. This aims to cover a blind spot of IOJankyIntervalsPerMinute as it cannot see an improvement where we unblock 19/20 threads performing I/O. Not reported across system sleep. More details @ https://bit.ly/chrome-io-jank-metric gab@chromium.org olivierli@chromium.org This metric is emitted every minute. It reports a number between 0 and 60 which corresponds to the number of 1 second intervals during the monitored minute to which a stalled foreground thread ScopedBlockingCall(BlockingType::MAY_BLOCK) was attributed. A stalled call is first attributed to the interval in which it began, and is attributed to a number of intervals equal to the length of the call, rounded to the nearest second. (Before http://crrev.com/c/2314873, the length was truncated rather than rounded.) This aims to capture moments where the main thread is responsive but the application is not (because things aren't loading). Not reported across system sleep. More details @ https://bit.ly/chrome-io-jank-metric robliao@chromium.org yuhengh@chromium.org Compositor presentation time when dragging tabs left or right to a different index. This metric is used to measure tab dragging performance. Record on the next compositor frame when the data is collected. robliao@chromium.org yuhengh@chromium.org Compositor presentation time when dragging tabs out to a new browser. This metric is used to measure tab dragging performance. Record on the next compositor frame when the data is collected. robliao@chromium.org yuhengh@chromium.org Compositor presentation time when dragging tabs. This metric is used to measure tab dragging performance. Record on the next compositor frame when the data is collected. robliao@chromium.org yuhengh@chromium.org Maxinum compositor presentation time when dragging tabs. This metric is used to measure tab dragging performance. Record after a drag and drop session is ended. jonross@chromium.org sky@chromium.org The time it takes for the browser to request a tab to be visible. Note that this only measures the time from the browser deciding to make the tab visible (after a click, or keyboard-shortcut to switch tabs), until the tab is requested to become visible. The time it takes to actually make the tab visible and show on screen is measured in a separate metric, in Browser.Tabs.TotalSwitchDuration3. This metric uses UMA_HISTOGRAM_CUSTOM_MICROSECONDS_TIMES so only users with high resolution clocks will report this metric. fdoray@chromium.org joenotcharles@google.com catan-team@chromium.org Whether a frame was presented after a tab switch event. Recorded based on the destination tab state when the tab switch starts: {TabSwitchingType} fdoray@chromium.org jonross@chromium.org joenotcharles@google.com catan-team@chromium.org The time between a tab switch input event timestamp and when the tab is hidden, if no frame is presented in the meantime. Recorded based on the destination tab state when the tab switch starts: {TabSwitchingType} jonross@chromium.org joenotcharles@google.com catan-team@chromium.org The time between the input event timestamp and the frame presentation time on a tab switch. Recorded based on the destination tab state when the tab switch starts: {TabSwitchingType} rhalavati@chromium.org chrome-privacy-core@google.com Number of open guest windows at the same time. Recorded when a new guest window is created. Please note that this metric double counts the lower numbers, meaning that if user opens three windows (without closing any in between), then buckets 1, 2, and 3 will all be incremented in turn. rhalavati@chromium.org chrome-privacy-core@google.com Number of open incognito windows at the same time. Recorded when a new incognito window is created. Please note that this metric double counts the lower numbers, meaning that if user opens three windows (without closing any in between), then buckets 1, 2, and 3 will all be incremented in turn. creis@chromium.org site-isolation-dev@chromium.org Recorded when RenderProcessHostImpl::FilterURL detects a disallowed URL sent by the renderer process and rewrites it to about:blank#blocked. wfh@chromium.org chrome-stability-core@google.com Count of renderer process terminations due to a code integrity failure, grouped by type. This is only reported on Windows. wfh@chromium.org chrome-stability-core@google.com Count of renderer process crashes grouped by type. wfh@chromium.org chrome-stability-core@google.com Count of renderer process kills grouped by type. oshima@chromium.org chrome-stability-core@google.com Out of BrowserRenderProcessHost.ChildKills, numer of kills due to SIGKILL, which is a strong signal of out of memory on Chrome OS, grouped by renderer type. wfh@chromium.org chrome-stability-core@google.com The launch error codes for failed renderer process launches. wfh@chromium.org chrome-stability-core@google.com Count of renderer process launch failures grouped by type. mych@chromium.org bfcache-dev@chromium.org Recorded whenever a RenderProcessHost is cleaned up, with the reason the cleanup fails or is delayed, or kNoDelay if it succeeded quickly. The total count of this histogram should be equal to BrowserRenderProcessHost.Cleanup.Total. mych@chromium.org bfcache-dev@chromium.org Records true when an attempt is made to clean up a RenderProcessHost. wfh@chromium.org chrome-stability-core@google.com Count of renderer process crashes that we miscounted because we took the exit code too early. Grouped by type. mych@chromium.org bfcache-dev@chromium.org Recorded whenever a RenderProcessHost is fast shutting down, with the reason the shutdown fails or is delayed, or kNoDelay if it succeeded quickly. The total count of this histogram should be equal to BrowserRenderProcessHost.FastShutdownIfPossible.Total. mych@chromium.org bfcache-dev@chromium.org Records true when an attempt is made to fast shutdown a RenderProcessHost. creis@chromium.org avi@chromium.org Reports false each time the Task Manager displays an unlabeled renderer process using the fallback task provider, indicating that a type of renderer process is missing a task provider, which is a bug. For comparison, this reports true when renderer processes are properly labeled by a primary task provider. These values are only reported when the Task Manager is open. This can be removed when no unlabeled cases remain, per https://crbug.com/739782. 4/2023; the data has been collected, verifying that less than 1% of routing ID space is used in practice. alexmos@chromium.org dcheng@chromium.org Record the percentage of routing IDs ever used by a RenderProcessHost when the C++ object is destroyed. The denominator is the range of all potentially usable routing IDs—usually std::numeric_limits<int32_t>::max()—but may be smaller (e.g. half that range). This histogram does not record data if the routing ID overflows; that event is captured as a browser crash instead. alexmos@chromium.org lukasza@chromium.org Records what happens when an attempt is made to use a spare RenderProcessHost - logs that either the attempt succeeded or why it failed. wfh@chromium.org chrome-stability-core@google.com Whether or not kernel-mode execution was recorded for a suspended renderer process. This is recorded after process creation but before process resumption. This is recorded for every sandboxed Renderer process launched on Windows only. wfh@chromium.org chrome-stability-core@google.com Whether or not user-mode execution was recorded for a suspended renderer process. This is recorded after process creation but before process resumption. This is recorded for every sandboxed Renderer process launched on Windows only. peconn@chromium.org peter@chromium.org Triggered when a TrustedWebActivity is paused, recording the time since it had been resumed. peconn@chromium.org peter@chromium.org The result of Digital Asset Link verification performed when starting a Trusted Web Activity. peconn@chromium.org peter@chromium.org Triggered when Trusted Web Activity verification completes. Records how long it takes the device to realise it is offline and fall back to the cached result. peconn@chromium.org peter@chromium.org Triggered when Trusted Web Activity verification completes. Records how long verification took when online. nicolaso@chromium.org pastarmovj@chromium.org Records which browser the Legacy Browser Support feature is configured to open (which defaults to IE). Recorded on browser startup, and when the value changes. Only measured when Legacy Browser Support is enabled via the BrowserSwitcher* policies. Desktop Chrome only. nicolaso@chromium.org pastarmovj@chromium.org Records whether the "AppData/Local/Google/BrowserSwitcher" directory was successfully created. Windows only. nicolaso@chromium.org pastarmovj@chromium.org Records whether a temp file in "AppData/Local/Google/BrowserSwitcher" the directory was successfully created. Windows only. nicolaso@chromium.org pastarmovj@chromium.org Records whether the temp file in "AppData/Local/Google/BrowserSwitcher" directory was successfully moved to its final destination ("cache.dat") after writing its contents. Windows only. nicolaso@chromium.org pastarmovj@chromium.org On navigation and redirect, whether to switch browsers or not. Only measured when Legacy Browser Support is enabled via the BrowserSwitcher* policies. When this is true, Chrome attempts to launch an alternative browser (typically IE) with the navigation URL. Desktop Chrome only. nicolaso@chromium.org pastarmovj@chromium.org On navigation and redirect, records how long it took to compute a decision (stay in Chrome/open in IE). Only measured when Legacy Browser Support is enabled via the BrowserSwitcher* policies. Desktop Chrome only. nicolaso@chromium.org pastarmovj@chromium.org Number of rules that come from an external greylist XML file. Measured when the sitelist is loaded during browser startup. Desktop Chrome only. nicolaso@chromium.org pastarmovj@chromium.org Number of rules that come from an external sitelist XML file. Measured when the sitelist is loaded during browser startup. Desktop Chrome only. nicolaso@chromium.org pastarmovj@chromium.org Number of rules in the BrowserSwitcherUrlGreylist policy. Measured when the policy is loaded. Desktop Chrome only. nicolaso@chromium.org pastarmovj@chromium.org Whether there is a wildcard ("*") rule in the BrowserSwitcherUrlGreylist policy. Measured when the BrowserSwitcherUrlGreylist policy gets updated. Desktop Chrome only. nicolaso@chromium.org pastarmovj@chromium.org Number of rules that come from Internet Explorer Enterprise Mode's sitelist. Measured when the sitelist is loaded during browser startup. Windows only. nicolaso@chromium.org pastarmovj@chromium.org Records whether an attempt to launch an alternate browser failed or succeeded. Desktop Chrome only. nicolaso@chromium.org pastarmovj@chromium.org Records how long it took to launch an alternate browser (regardless of whether it failed or succeeded). Desktop Chrome only. nicolaso@chromium.org pastarmovj@chromium.org Number of rules in the BrowserSwitcherUrlList policy. Measured when the policy gets updated. Desktop Chrome only. nicolaso@chromium.org pastarmovj@chromium.org Whether there is a wildcard ("*") rule in the BrowserSwitcherUrlList policy. Measured when the BrowserSwitcherUrlList policy gets updated. Desktop Chrome only. swestphal@chromium.org rayankans@chromium.org Records the number of fingerprints in assetlinks.json for every performed Digital Asset Link validation with a valid `sha256_cert_fingerprints`-entry. rkaplow@chromium.org src/components/infobars/OWNERS Records how many times a particular infobar was shown to the user. On Android, the infobar may be partially hidden due to stacking. On desktop, a tab with a visible infobar will increment this histogram every time the user switches away from the tab and back to it with the infobar still visible. mdjones@chromium.org Records how many times a particular infobar was displayed in a hidden state due to at least one other infobar existing. If there are no existing infobars, but the user scrolls at the exact moment a new infobar is added, the infobar may be hidden, but not recorded in this metric. Reliably triggering and detecting this combination is tricky so it isn't done. Only recorded on Android. mdjones@chromium.org Records how many times a particular infobar was visible and hiding a new infobar behind it. Only recorded on Android. mdjones@chromium.org Records how many times a particular infobar was displayed in a visible state due to no other infobars existing. If there are no existing infobars, but the user scrolls at the exact moment a new infobar is added, the infobar may be hidden, but recorded here as visible. Reliably triggering and detecting this combination is tricky so it isn't done. Only recorded on Android. corising@chromium.org chrome-desktop-ui-sea@google.com Records true when the side panel combobox is opened. corising@chromium.org chrome-desktop-ui-sea@google.com Records how long the side panel was open for (capped at 1 hour) starting when the side panel is triggered to be opened. Note this might be different than when the side panel becomes visible due to delays for loading content. Recorded when the side panel is closed. corising@chromium.org chrome-desktop-ui-sea@google.com Logs the UI location from which the side panel is triggered to be opened or to change the entry shown. Recorded when the side panel is shown or the entry changed and a trigger is provided. corising@chromium.org chrome-desktop-ui-sea@google.com Logs the UI location from which the side panel is triggered to be opened. Recorded when the side panel is shown and a trigger is provided. emshack@chromium.org chrome-desktop-ui-sea@google.com Records the width of the side panel contents after the side panel was resized. emshack@chromium.org chrome-desktop-ui-sea@google.com Records the width of the side panel contents as a percentage of the browser window width after the side panel was resized. emshack@chromium.org chrome-desktop-ui-sea@google.com Records the width of the side panel contents after the side panel was resized, while the {SidePanelEntry} side panel entry was displayed. emshack@chromium.org chrome-desktop-ui-sea@google.com Records the width of the side panel contents as a percentage of the browser window width after the side panel was resized, while the {SidePanelEntry} side panel entry was displayed. corising@chromium.org chrome-desktop-ui-sea@google.com Records how long the {SidePanelEntry} side panel entry was open for (capped at 1 hour). This is recorded in SidePanelEntry::OnEntryHidden when the entry is hidden (ie. switching entries or the side panel closing). corising@chromium.org chrome-desktop-ui-sea@google.com Records how a side panel entry is triggered to be shown. (i.e. by toolbar button, by commbobox selection). This is recorded when a new side panel entry is triggered to be shown regardless any side panel entry is currently open. corising@chromium.org chrome-desktop-ui-sea@google.com Records elapsed time between when the {SidePanelEntry} side panel entry triggered to be shown and when it was shown.