caitlinfischer@google.com src/base/metrics/OWNERS The size (up to 1M bytes) of the variations seed (serialized uncompressed proto) that is loaded on startup and used. Note that not all loaded seeds are used; for example, expired seeds are not used. The seed could be a regular seed or a safe seed. Recorded after setting up field trials with the seed. caitlinfischer@google.com src/base/metrics/OWNERS The size (up to 10M bytes) of the variations seed (serialized uncompressed proto) that is loaded on startup and used. Note that not all loaded seeds are used; for example, expired seeds are not used. The seed could be a regular seed or a safe seed. Recorded after setting up field trials with the seed. Windows only. caitlinfischer@google.com src/base/metrics/OWNERS The number of studies contained in the seed that is being used. Note that the client may not participate in all studies due to study filtering. The seed could be a regular seed or a safe seed. Recorded before setting up field trials with the seed. lucnguyen@google.com rogerm@google.com asvitkine@chromium.org src/base/metrics/OWNERS The result of attempting to apply a runtime mutable study from a new variations seed. Recorded for each runtime mutable study in the seed whenever a new seed is successfully stored. ntfschr@chromium.org src/android_webview/OWNERS The time interval between when a WebView-using app's copy of the variations seed was written to its local app storage and when it was loaded for use. This is different from Variations.SeedFreshness because it measures freshness from when the app received its copy of the seed, not when the seed was originally downloaded. Written every time the browser process initializes, but not if there was an error reading the seed, or no seed existed. lucnguyen@google.com src/base/metrics/OWNERS Whether creating a simulated field trial for a variations study results in the same assigned group and nullability state as creating the actual field trial for the study. Recorded for approximately 1% of studies when creating trials from a variations seed at startup, excluding studies that already exist or have feature conflicts. caitlinfischer@google.com src/base/metrics/OWNERS 1456342 The error code given by JSONFileValueDeserializer::Deserialize(). Emitted during CleanExitBeacon initialization when the beacon file cannot be deserialized. In other words, it is emitted whenever the "Not deserializable" bucket of the BeaconFileStateAtStartup histogram is emitted. As of M102 for desktop and iOS and M103 for Android Chrome, any client may emit this metric. Prior to this, the metric was emitted by only clients in an enabled group that used the beacon file. caitlinfischer@google.com src/base/metrics/OWNERS 1456342 The state of the beacon file and its contents, e.g. readable or not deserializable. See the BeaconFileDeserializationError histogram for more details on why the file was not deserializable. Emitted when the CleanExitBeacon is initialized. As of M102 for desktop and iOS and M103 for Android Chrome, any client may emit this metric. Prior to this, the metric was emitted by only clients in an enabled group that used the beacon file. caitlinfischer@google.com src/base/metrics/OWNERS 1456342 Denotes whether the beacon file was successfully written. Recorded immediately after Chrome tries to write the file. As of M102 for desktop and iOS and M103 for Android Chrome, any client may emit this metric. Prior to this, the metric was emitted by only clients in an enabled group that used the beacon file. Expired 02/2025. Revived M139. artomczak@google.com src/base/metrics/OWNERS Records the hash of the feature name for every feature accessed via base::FeatureList::IsEnabled(). This is recorded once per feature per session to avoid performance overhead. artomczak@google.com src/base/metrics/OWNERS Records the hash of the feature name for every feature accessed via base::FeatureList::IsEnabled() *before* the FeatureList has been initialized. This is recorded once per feature per session. thiabaud@google.com src/base/metrics/OWNERS Logged whenever |base::LogInvalidValue| is called, i.e. whenever an invalid value is used with a FeatureParam. This is normally due to a bad config file. The logged value of this histogram is a hash of the study name associated with the aforementioned feature. asvitkine@chromium.org rkaplow@chromium.org swestphal@chromium.org src/android_webview/OWNERS src/base/metrics/OWNERS The state of the delta-compression when attempting to fetch a variations seed via the java-based fetcher on Android. Records both the requested and received delta compression values in one enumeration. asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS The latency of connection to the variations server when fetching an initial variations seed during Android Chrome first run. This is included in the Variations.FirstRun.SeedFetchTime measurement. Only considers cases where an HTTP 200 result was received. Note: This expired in January 2025 and was revived in M143. asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS The result of attempting to fetch an initial variations seed during Android Chrome first run. Records both the HTTP code and various error values in one enumeration. Prior to M95, non-200 HTTP codes were counted as IOException. Starting with M95, these codes are logged separately. asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS The latency of fetching an initial variations seed during Android Chrome first run. Only considers cases where an HTTP 200 result was received. asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS The result of attempting to import an initial variations seed during Android and iOS Chrome first run. Logged from VariationsSeedStore::LoadSeed when seed prefs do not exist. For non-first run launches, this value is expected to be "failed - no first run seed." Supported for iOS Chrome since M135. harringtond@google.com src/base/metrics/OWNERS Reports whether any field trials were overridden with the metrics-internals field trials tab. This is recorded once at startup. caitlinfischer@google.com src/base/metrics/OWNERS The owner of the top-level domain from which certain subframe-initiated HTTP requests are made. It's logged after determining that (i) the request should include variations headers, (ii) the request is from a subframe, and (iii) TrustedParams are not present. It's logged before adding headers. It allows us to verify whether non-render-thread-initiated requests, if any, lack TrustedParams. asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS Records number of experiment ids in the X-Client-Data header at the time the header is constructed. asvitkine@chromium.org caitlinfischer@google.com src/base/metrics/OWNERS Details about the request context in which an HTTP request is made. Logged after determining that the request should include variations headers but before adding the headers. rogerm@google.com src/base/metrics/OWNERS The time source used by the variations service to get the current time. asvitkine@chromium.org src/base/metrics/OWNERS Records the reason for rejecting an invalid layer. Recorded when processing the variations seed at startup and when simulating a new seed. asvitkine@chromium.org src/base/metrics/OWNERS Records the reason for rejecting an invalid study. Recorded when processing the variations seed at startup and when simulating a new seed. yulunz@chromium.org src/base/metrics/OWNERS Records the reason for rejecting a variations seed that contains a layer with LIMITED entropy mode. This is recorded when calculating the entropy usage for the studies constrained to the layer with LIMITED entropy mode. caitlinfischer@google.com src/base/metrics/OWNERS The number of kilobytes of the variations crash keys, recorded every time crash keys are updated. Crash keys larger than kVariationKeySize will be truncated. The size of kVariationKeySize is controlled by the preprocessor macro LARGE_VARIATION_KEY_SIZE. {Size}; asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS Records how the country code saved in prefs used for filtering permanent consistency studies compares to the country code in the variations seed. This is recorded each time the saved country code is loaded from the pref. asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS The result of verifying the latest variations seed's signature, recorded when the seed is loaded from Local State. Not recorded when running in safe mode. ydago@chromium.org pastarmovj@chromium.org src/base/metrics/OWNERS Records the restrictions on how experiment variations are applied on the client. This is recorded at start up. This histogram is not interesting anymore but it is still used by test code to verify that the respective log is properly applied so it will be kept in the code even if not collected by the server anymore. TODO(crbug/1345624): Clean this dependency up and remove the histogram. Kept as a diagnostic metric. asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS Counts the number of times the VariationsService is allowed or not allowed to make a request due to the ResourceRequestAllowedNotifier. Useful for debugging cases where variations seeds may not be getting fetched. rogerm@google.com src/base/metrics/OWNERS Captures the hashed name of a feature for which a runtime mutability operation (registration, state query, state update, etc) failed. rogerm@google.com src/base/metrics/OWNERS Captures the result of applying a runtime mutability update. Recored once per runtime-mutable feature update received in each downloaded variations seed. caitlinfischer@google.com src/base/metrics/OWNERS The absolute value of the difference between the beacon-file-based crash streak and the local-state-based crash streak. Recorded during startup after successfully reading the beacon file's crash streak and before any updates are made (in memory) to the Local State crash streak. Note: This expired in May 2025 and was revived in M143. asvitkine@chromium.org src/base/metrics/OWNERS Records whether the safe variations seed was successfully read from disk on startup. Records a detailed reason on read failure. Only recorded when attempting to run in safe mode. Expired after M77 and revived in M93. asvitkine@chromium.org src/base/metrics/OWNERS The result of verifying the safe variations seed's signature. Recorded when the seed is loaded from disk. Only recorded when attempting to run in safe mode. Expired after M77 and revived in M93. asvitkine@chromium.org src/base/metrics/OWNERS Records the result of attempting to store a safe variations seed (and all associated metadata) on disk. Expired after M77 and revived in M93. asvitkine@chromium.org src/base/metrics/OWNERS The result of verifying the safe variations seed's signature. Recorded when attempting to store the safe seed on disk. Expired after M77 and revived in M93. asvitkine@chromium.org src/base/metrics/OWNERS The number of consecutive crashes observed by the VariationsService without a single intervening successful seed fetch. Recorded during Chrome startup when the MetricsStateManager's CleanExitBeacon is created. Prior to M92, this was recorded when the SafeSeedManager was created. asvitkine@chromium.org src/base/metrics/OWNERS The number of consecutive failed attempts to fetch a new seed by the VariationsService. Recorded during Chrome startup when the SafeSeedManager is created. hnakashima@chromium.org src/chrome/browser/flags/OWNERS The number of consecutive crashes observed by Android's CachedFeatureFlags without a single intervening successful flag cache from native. Recorded the first time a cached flag is checked. Android only. hnakashima@chromium.org src/chrome/browser/flags/OWNERS What behavior of Safe Mode for CachedFlags (including not engaging it) was performed. {Event} The ratio "Cached" / ("WillCache" - "Pause") is the success rate for a given mode, and (100% - success rate) is the crash rate. Android only. jwd@chromium.org asvitkine@chromium.org src/base/metrics/OWNERS Counts if a response from the variations server is the first response of the day or not. This is counted when a new valid seed or a 304 is received. The date line is computed in UTC and the times being compared are the server time from the server response and the stored server time from the last successful request. asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS HTTP response codes and network error encountered by VariationsService when attempting to fetch a variations seed from the server over an {ConnectionType} connection. ginnyhuang@chromium.org src/base/metrics/OWNERS The latency of fetching an initial variations seed during first run on non-Android platforms. Only considers cases where an HTTP 200 result was received. The respective metric for Android's is Variations.FirstRun.SeedFetchTime. Note: This expired in May 2025 and was revived in M143. rcanoaparicio@google.com src/base/metrics/OWNERS The time it takes to compress the {SeedFile} info before writing it to disk. Recorded on a background thread every time a seed is written to disk. rcanoaparicio@google.com src/base/metrics/OWNERS The time it takes to decompress the {SeedFile} info after reading it from disk. Recorded on a background thread every time a seed is read from disk. caitlinfischer@google.com src/base/metrics/OWNERS Records the success or failure of reading a seed file, distinguishing between safe seed file and latest seed file. This metric is emitted only by clients in the seed file treatment group. {SeedFile} rcanoaparicio@google.com src/base/metrics/OWNERS Records the result of reading the seed file. This metric is emitted only by clients in the seed file treatment group. caitlinfischer@google.com src/base/metrics/OWNERS Whether a seed copied from local state to the {SeedFile} is empty. Emitted by clients in the seed file experiment's treatment group when creating the {SeedFile}'s SeedReaderWriter if the file (a) does not exist or (b) exists but cannot be read. asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS The time interval between when the variations seed was downloaded and when it was loaded for use. Recorded after the seed is loaded for both regular seeds and safe seeds. Not recorded for expired seeds or for seeds with an unknown download time, which is typically a seed imported via first run. A seed that (a) was fetched over 30 days ago and (b) is older than the binary build time is considered expired. Prior to M93, this metric was recorded under slightly different conditions. First, the metric was recorded for safe seeds--expired or not--as long as they had download times. As of M93, it's no longer recorded for expired safe seeds. Second, the metric was not previously recorded for regular seeds that were older than 30 days, but as of M93, the freshness is recorded for seeds that are older than 30 days and newer than the binary build time. Prior to M99, this histogram would not be logged for first run seeds (where they were not fetched by Chrome, but provided externally, such as part of Chrome installation via Google Update). In M99+, this histogram will be logged with a value of 0 minutes, so that the seed is considered fresh. ntfschr@chromium.org src/android_webview/OWNERS Logs the difference between the age of the seed (Seed Freshness) and the app's copy of the seed (App Seed Freshness). Seed Freshness is recorded when seed is loaded for both regular and safe seeds and AppSeed Freshness is recorded every time the browser process initializes. SeedFreshnessDiff is recorded at the latter of these two events. In the case when a user's clock is set incorrectly, or when a user changes their clock backwards in between Chrome runs, the difference between the seed freshness and the app seed freshness will be calculated and cached. ntfschr@chromium.org src/base/metrics/OWNERS src/android_webview/OWNERS Records the time spent blocking WebView startup to wait for the variations seed. This is less than the entire time needed to load the seed, since startup is only blocked if loading is not complete by the time the seed is needed. This would ideally have a maximum value equal to the timeout in VariationsSeedLoader, but there is no actual limit on how late that timeout could fire. This is logged once per WebView startup, whether or not loading the seed was successful. asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS Records whether the latest variations seed was successfully read from local state on startup. Records a detailed reason on read failure. Not recorded when running in safe mode. asvitkine@chromium.org src/base/metrics/OWNERS Records how long it takes to load and apply a variations seed. Recorded only when the seed is successfully loaded and applied. To help diagnose a processing time increase, see whether Variations.AppliedSeed.Size and Variations.AppliedSeed.StudyCount increased at the same time. rcanoaparicio@google.com src/base/metrics/OWNERS Records the source where the seed was read from successfully. asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS Records the result of storing the variations seed that was received from the server. Also logs the types of data received (gzip, delta, etc). caitlinfischer@google.com src/base/metrics/OWNERS Captures details about Chrome's attempt to use a VariationsSeed, e.g. whether Chrome used a safe seed. Recorded once during Chrome startup after attempting to load a seed and before field trials are created. This metric is not accurate between M123 and M141, inclusively. In these milestones, the buckets for applying regular or safe seeds may be emitted to even though the seed is later dropped due to misconfiguration. asvitkine@chromium.org src/base/metrics/OWNERS The hash of the study name of a sticky activation study when it is activated for the first time (i.e. not sticky from a previous session). Logged for verification purposes to verify that studies stay sticky. asvitkine@chromium.org src/base/metrics/OWNERS The hash of the study name of a sticky activation study when it is parsed from the pref value. Logged for verification purposes to verify that studies stay sticky. asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS The size of the variations seed data, which may be a partial (delta) or compressed (gzip) payload, see Variations.SeedStoreResult. asvitkine@chromium.org src/base/metrics/OWNERS The run time of the VariationsSeedStore::StoreSeedData function until the callback is called with the result. Changed in M145. asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS The result of verifying the variations seed signature, recorded when the variations seed is stored to Local State after being retrieved from the server. asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS The time since the previous attempt to fetch the variations seed within the same session, with 0 indicating that this is the first attempt. Recorded when a variations seed fetch is attempted by the VariationsService. asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS Log the user channel assigned at browser startup used for evaluating the variations seeds. svenzheng@google.com vivianz@google.com Capture the status of variations state encryption in Chrome feedback report. ntfschr@chromium.org src/android_webview/OWNERS The time between the start of two consecutive WebView seed downloads by WebView's service. This metric will not be written if seeds are never downloaded. ntfschr@chromium.org src/android_webview/OWNERS The delay between when a WebView seed download was scheduled by the service with the JobScheduler and when it was executed. asvitkine@chromium.org rkaplow@chromium.org src/base/metrics/OWNERS Stores the state of the loaded variations seed, e.g. not expired. Recorded after the seed is loaded and before trials are created from the seed. Note that seeds for which this metric is recorded may not be applied, e.g. when the seed has expired. {Mode}