| # Copyright 2019 The Fuchsia Authors. All rights reserved. |
| # Use of this source code is governed by a BSD-style license that can be |
| # found in the LICENSE file. |
| |
| ################################################################################ |
| # Cobalt Project: local_storage |
| ################################################################################ |
| |
| metric_definitions: |
| |
| ################################################################################ |
| # Corruption Event |
| # |
| # A corruption event logs a '1' every time a component detects corruption. This |
| # can be either data or metadata corruption depending on the source. |
| ################################################################################ |
| - id: 48 |
| metric_name: corruption_events_migrated |
| # This will always log a single event (1). |
| metric_type: OCCURRENCE |
| metric_semantics: [USAGE_COUNTING] |
| metric_dimensions: |
| - dimension: "source" |
| event_codes: |
| 0: Unknown |
| 1: Fvm |
| 2: BlobFs |
| 3: MinFs |
| 4: FxFs |
| - dimension: "corruption_type" |
| event_codes: |
| 0: Unknown |
| 1: Data |
| 2: Metadata |
| reports: |
| ############################################################################ |
| # A fleet-wide summary. |
| ############################################################################ |
| - report_name: corruption_per_device |
| id: 2 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Compression format |
| # |
| # Total file sizes (in bytes) per compression format. Calculated on mount. |
| # |
| # Currently only supported by blobfs. When blobfs is starting up after boot, we |
| # compute the total (uncompressed) sizes of all blobs per compression format, |
| # and log the counts to Cobalt. |
| # |
| ################################################################################ |
| - id: 49 |
| metric_name: compression_format_migrated |
| metric_type: INTEGER |
| metric_units: BYTES |
| metric_semantics: [DATA_SIZE] |
| metric_dimensions: |
| - dimension: "source" |
| event_codes: |
| 0: Unknown |
| 1: BlobFs |
| - dimension: "format" |
| event_codes: |
| 0: Unknown |
| 1: Uncompressed |
| 2: LZ4 |
| 3: ZSTD |
| 4: ZSTDSeekable |
| 5: ZSTDChunked |
| reports: |
| ############################################################################ |
| # A per-device histogram report. |
| ############################################################################ |
| - report_name: compression_per_device_histogram |
| id: 1 |
| report_type: UNIQUE_DEVICE_HISTOGRAMS |
| system_profile_selection: SELECT_LAST |
| ########################################################################## |
| # Since we only log the count once per boot, we use MAX here in case a |
| # device reboots several times during the day. |
| ########################################################################## |
| local_aggregation_procedure: MAX_PROCEDURE |
| int_buckets: |
| linear: |
| step_size: 20971520 # 20MB |
| num_buckets: 100 |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Page fault latency |
| # |
| # Histogram of latencies in nanoseconds for satisfying a page fault. |
| # |
| # Each time a page fault occurs (a client accesses a not-present page in a |
| # pager-backed VMO), the latency of the userspace fault handling is recorded |
| # and its latency (in nanoseconds) is added to the histogram. |
| # |
| # Blobfs currently flushes this metric to Cobalt every five minutes (at the |
| # same time all other metrics are flushed.) |
| # |
| # Currently only supported by blobfs, and only when blobfs.userpager=true is set |
| # in the build. |
| # |
| ################################################################################ |
| - id: 50 |
| metric_name: page_fault_latency_migrated |
| metric_type: INTEGER_HISTOGRAM |
| metric_units: NANOSECONDS |
| metric_semantics: [LATENCY] |
| int_buckets: |
| exponential: |
| floor: 0 |
| num_buckets: 15 |
| initial_step: 10000 |
| step_multiplier: 2 |
| metric_dimensions: |
| - dimension: "source" |
| event_codes: |
| 0: Unknown |
| 1: BlobFs |
| - dimension: "format" |
| event_codes: |
| 0: Unknown |
| 1: Uncompressed |
| 2: LZ4 |
| 3: ZSTD |
| 4: ZSTDSeekable |
| 5: ZSTDChunked |
| - dimension: "read_ahead_size" |
| event_codes: |
| 0: 8KiB |
| 1: 16KiB |
| 2: 32KiB |
| 3: 64KiB |
| 4: 128KiB |
| 5: 256KiB |
| 6: 512KiB |
| 7: LargerThan512KiB |
| reports: |
| ############################################################################ |
| # A fleet-wide histogram report |
| ############################################################################ |
| - report_name: page_fault_latency_histogram |
| id: 1 |
| report_type: FLEETWIDE_HISTOGRAMS |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # On-disk version |
| # |
| # The component string field tracks the version in one of two formats: |
| # * `${major}/${oldest_rev}` for Blobfs, Minfs |
| # * `${major}.${minor}` for Fxfs |
| # A dimension is used to track the real storage sub-component. The enumeration of possible version |
| # number combinations is tracked by versions.txt. This metric is incremented upon every mount. |
| ################################################################################ |
| - id: 61 |
| metric_name: version_migrated |
| metric_type: STRING |
| string_candidate_file: fuchsia/local_storage/versions.txt |
| metric_semantics: [USAGE_COUNTING] |
| metric_dimensions: |
| - &source_dimension |
| dimension: source |
| event_codes: |
| 0: unknown |
| 1: fvm |
| 2: blobfs |
| 3: minfs |
| 4: fxfs |
| reports: |
| - report_name: version_mount_counts |
| id: 1 |
| report_type: STRING_COUNTS |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| string_buffer_max: 10 |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ##################################################################### |
| # Total inodes |
| # |
| # Calculated on mount. |
| # |
| # Total number of inodes, either free or in use, in the system. |
| # The metrics will help us understand the size of inode table. This number in |
| # might also help understand other fragmentation metrics, namely file_in_use and |
| # extent_containers_in_use better. |
| ##################################################################### |
| - id: 62 |
| metric_name: total_inodes_migrated |
| metric_type: INTEGER |
| metric_dimensions: &fragmentation_dimensions |
| - *source_dimension |
| metric_units_other: "inodes" |
| metric_semantics: [MEMORY_USAGE] |
| reports: |
| - report_name: total_inodes_count |
| id: 1 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: MAX_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ##################################################################### |
| # Files in use |
| # |
| # Calculated on mount. |
| # |
| # Total number of files, (or blobs in case of blobfs), in use in the system. |
| # This number may vary largely across builds or when the device has additional |
| # packages installed other than system base packages. |
| ##################################################################### |
| - id: 63 |
| metric_name: file_in_use_migrated |
| metric_type: INTEGER |
| metric_dimensions: *fragmentation_dimensions |
| metric_units_other: "files" |
| metric_semantics: [MEMORY_USAGE] |
| reports: |
| - report_name: files_in_use_count |
| id: 1 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: MAX_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| |
| ##################################################################### |
| # Extent containers in use |
| # |
| # Calculated on mount. |
| # |
| # File/blob may use one or more extents, a pointer to a set of blocks. Extent |
| # containers are collection of one or more extents. Large the number of extent |
| # containers in use indicate higher fragmentation of the storage. |
| ##################################################################### |
| - id: 64 |
| metric_name: extent_containers_in_use_migrated |
| metric_type: INTEGER |
| metric_dimensions: *fragmentation_dimensions |
| metric_units_other: "extent containers" |
| metric_semantics: [MEMORY_USAGE] |
| reports: |
| - report_name: extent_containers_in_use_count |
| id: 1 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: MAX_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Extents per file fragmentation |
| # |
| # A histogram of extents used by a file/blob calculated on mount. |
| # |
| # The histogram presents fragmentation at file/blob level. This metric will help |
| # in understanding how defragmentation of certain files will help either to |
| # recover extents or to improve read performance. |
| ################################################################################ |
| - id: 65 |
| metric_name: extents_per_file_migrated |
| metric_type: INTEGER_HISTOGRAM |
| metric_dimensions: *fragmentation_dimensions |
| metric_units_other: "extent containers" |
| metric_semantics: [MEMORY_USAGE] |
| int_buckets: |
| exponential: |
| floor: 0 |
| num_buckets: 10 |
| initial_step: 10 |
| step_multiplier: 2 |
| reports: |
| ############################################################################ |
| # A fleet-wide histogram report |
| ############################################################################ |
| - report_name: fragmentation_extents_per_file_histogram |
| id: 1 |
| report_type: FLEETWIDE_HISTOGRAMS |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # In use block fragmentation |
| # |
| # A histogram of used contiguous blocks (representing contiguous offsets of a |
| # file) calculated on mount. |
| # |
| # The histogram shows used block fragmentation. This metric will help |
| # in understand the effects of block allocation policy. The metric will also |
| # help in targeting what blocks to move if we decide to defragment the system. |
| ################################################################################ |
| - id: 66 |
| metric_name: in_use_fragments_migrated |
| metric_type: INTEGER_HISTOGRAM |
| metric_dimensions: *fragmentation_dimensions |
| metric_units_other: "fragments" |
| metric_semantics: [MEMORY_USAGE] |
| int_buckets: |
| exponential: |
| floor: 0 |
| num_buckets: 10 |
| initial_step: 10 |
| step_multiplier: 2 |
| reports: |
| ############################################################################ |
| # A fleet-wide histogram report |
| ############################################################################ |
| - report_name: fragmentation_in_use_fragments_histogram |
| id: 1 |
| report_type: FLEETWIDE_HISTOGRAMS |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Free fragments. |
| # |
| # A histogram of contiguous free blocks calculated on each mount. |
| # |
| # The histogram shows free block fragmentation. This metric will help |
| # in understand likelyhood of running out of extents during a OTA. |
| ################################################################################ |
| - id: 67 |
| metric_name: free_fragments_migrated |
| metric_type: INTEGER_HISTOGRAM |
| metric_dimensions: *fragmentation_dimensions |
| metric_units_other: "fragments" |
| metric_semantics: [MEMORY_USAGE] |
| int_buckets: |
| exponential: |
| floor: 0 |
| num_buckets: 10 |
| initial_step: 10 |
| step_multiplier: 2 |
| reports: |
| ############################################################################ |
| # A fleet-wide histogram report |
| ############################################################################ |
| - report_name: fragmentation_free_fragments_histogram |
| id: 1 |
| report_type: FLEETWIDE_HISTOGRAMS |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| |
| ################################################################################ |
| # Vnode Operation Latency |
| # |
| # A histogram of the latency of Vnode-level operations, in microseconds, logged |
| # every few minutes. This is a replacement for metrics 35-47. |
| ################################################################################ |
| - id: 70 |
| metric_name: vnode_operation_latency |
| metric_type: INTEGER_HISTOGRAM |
| metric_dimensions: &vnode_op_dimensions |
| - *source_dimension |
| - dimension: operation |
| event_codes: |
| 0: unknown |
| 1: append |
| 2: close |
| 3: create |
| 4: get_attr |
| 5: link |
| 6: lookup |
| 7: read |
| 8: read_dir |
| 9: set_attr |
| 10: sync |
| 11: truncate |
| 12: unlink |
| 13: write |
| metric_units: MICROSECONDS |
| metric_semantics: [LATENCY] |
| int_buckets: |
| exponential: |
| floor: 0 |
| initial_step: 5 |
| step_multiplier: 2 |
| num_buckets: 16 |
| reports: |
| ############################################################################ |
| # A fleet-wide histogram report |
| ############################################################################ |
| - report_name: histogram |
| id: 1 |
| report_type: FLEETWIDE_HISTOGRAMS |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| |
| # Next ID: 71 |
| |
| |
| ################################################################################ |
| |
| # FTL and Nand Metrics |
| # Reserved Range: [2000, 3000) |
| |
| ################################################################################ |
| # FTL: Wear Count |
| # |
| # A counter representing the highest wear count of the FTL driver. |
| # |
| # A NAND block has an expected lifetime in erase cycles. An erase must occur |
| # before anything can be written (updating data). |
| # |
| # This count is the highest number of erase cycles per block, across all |
| # erase blocks in the NAND device. |
| # |
| # This metric allows observing how the FTL wear leveling algorithm is performing, |
| # such that the device lifetime is maximized, because the higher the wear the |
| # more probable the block is expected to fail. |
| # |
| # Sampling Frequency: 1 sample per hour. |
| ################################################################################ |
| - id: 2003 |
| metric_name: ftl_wear_count_migrated |
| metric_type: INTEGER |
| metric_units_other: wear count |
| metric_semantics: [USAGE_COUNTING] |
| reports: |
| ############################################################################ |
| # A fleet-wide summary. |
| ############################################################################ |
| - report_name: ftl_wear_count_per_device |
| id: 1 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: MAX_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| - report_name: ftl_wear_count_histogram_per_device |
| id: 2 |
| report_type: FLEETWIDE_HISTOGRAMS |
| int_buckets: |
| exponential: |
| floor: 0 |
| num_buckets: 20 |
| initial_step: 1 |
| step_multiplier: 2 |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # FTL: Block Operation Count |
| # |
| # A counter representing the number of block operations the FTL receives. |
| # |
| # Allows measuring the IO pressure on the FTL. When combined with the |
| # |ftl_nand_operation_count| an approximate operation amplification |
| # can be estimated. |
| # |
| # Sampling Frequency: 1 sample per hour. |
| # |
| # |operation_type|: Refers to the type of block operation received in the FTL Layer. |
| # |
| ################################################################################ |
| - id: 2004 |
| metric_name: ftl_block_operation_count_migrated |
| metric_type: INTEGER |
| metric_semantics: [USAGE_COUNTING] |
| metric_units_other: "block operations" |
| metric_dimensions: |
| - dimension: operation_type |
| event_codes: |
| 0: Unknown |
| 1: BLOCK_READ |
| 2: BLOCK_WRITE |
| 3: BLOCK_FLUSH |
| 4: BLOCK_TRIM |
| reports: |
| ############################################################################ |
| # A fleet-wide summary. |
| ############################################################################ |
| - report_name: ftl_block_operation_per_device |
| id: 1 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: SUM_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| - report_name: ftl_block_max_operations_per_device |
| id: 2 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: MAX_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| - report_name: ftl_block_operation_histogram_per_device |
| id: 3 |
| report_type: FLEETWIDE_HISTOGRAMS |
| int_buckets: |
| exponential: |
| floor: 0 |
| num_buckets: 20 |
| initial_step: 1 |
| step_multiplier: 2 |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # FTL: NAND Operation Count |
| # |
| # A counter representing the number of block operations the FTL receives. |
| # |
| # Allows measuring the IO pressure on the FTL. When combined with the |
| # |ftl_nand_operation_count| an approximate operation amplification |
| # can be estimated. |
| # |
| # Sampling Frequency: 1 sample per hour. |
| # |
| # |operation_type|: Refers to the type of NAND operation issued to the NAND layer by the ftl, |
| # in response to a given operation type. |
| # |source_operation_type|: Refers to the type of block operation received in the FTL Layer. |
| # |
| # |
| ################################################################################ |
| - id: 2005 |
| metric_name: ftl_nand_operation_count_migrated |
| metric_type: INTEGER |
| metric_units_other: "nand operations" |
| metric_semantics: [USAGE_COUNTING] |
| metric_dimensions: |
| - dimension: operation_type |
| event_codes: |
| 0: Unknown |
| 1: NAND_READ |
| 2: NAND_WRITE |
| 3: NAND_ERASE |
| - dimension: source_operation_type |
| event_codes: |
| 0: Unknown |
| 1: BLOCK_READ |
| 2: BLOCK_WRITE |
| 3: BLOCK_FLUSH |
| 4: BLOCK_TRIM |
| reports: |
| ############################################################################ |
| # A fleet-wide summary. |
| ############################################################################ |
| - report_name: ftl_nand_operations_per_device |
| id: 1 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: SUM_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| - report_name: ftl_nand_operation_max_per_device |
| id: 2 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: MAX_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| - report_name: ftl_nand_operation_histogram_per_device |
| id: 3 |
| report_type: FLEETWIDE_HISTOGRAMS |
| int_buckets: |
| exponential: |
| floor: 0 |
| num_buckets: 20 |
| initial_step: 1 |
| step_multiplier: 2 |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Nand bit flips corrected by ECC per call. |
| # |
| # A histogram of the number of bitflips correct by ECC on a read. Excess bit |
| # flips that cannot be corrected will be set as the max number of correctable |
| # plus one. For all current use cases, 30 is the max correctable bits. |
| # |
| # Provides insight into the frequency and extremity of bit flips observed. |
| ################################################################################ |
| - id: 2006 |
| metric_name: nand_read_ecc_bit_flips |
| metric_type: INTEGER_HISTOGRAM |
| metric_semantics: [USAGE_COUNTING] |
| metric_units_other: "bitflips" |
| int_buckets: |
| linear: |
| # 0, 1, 2... 29, 30, 31 |
| step_size: 1 |
| num_buckets: 32 |
| reports: |
| ############################################################################ |
| # A fleet-wide histogram of bit flips on each read. |
| ############################################################################ |
| - report_name: nand_ecc_bit_flips_histogram |
| id: 1 |
| report_type: FLEETWIDE_HISTOGRAMS |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Repeated read attempts to nand before a successful read, or giving up. |
| # |
| # A histogram of the number of read attempts used on a single read, always at |
| # least 1 for succeeding on the first attempt, and if it gives up it reports |
| # MAX_ULONG and ends up in the overflow bucket. |
| # |
| # For insight into how many consecutive read attempts are actually productive, |
| # and how much benefit we get from our current maximum. |
| ################################################################################ |
| - id: 2007 |
| metric_name: nand_read_attempts_per_read |
| metric_type: INTEGER_HISTOGRAM |
| metric_semantics: [USAGE_COUNTING] |
| metric_units_other: "attempts" |
| int_buckets: |
| exponential: |
| # 0, 1, 2, 4, 8... 64, 128 |
| floor: 0 |
| num_buckets: 9 |
| initial_step: 1 |
| step_multiplier: 2 |
| reports: |
| ############################################################################ |
| # A fleet-wide histogram of the number of attempts required to read. |
| ############################################################################ |
| - report_name: attempts_histogram |
| id: 1 |
| report_type: FLEETWIDE_HISTOGRAMS |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Number of failed read attempts to the nand. |
| # |
| # A count of how many read attempts are sent to the nand which result in |
| # failure. |
| # |
| # For insight into how fleetwide failures might be clustered in a subset of |
| # devices, and how bad some of those devices might be. Expecting daily values |
| # to be normally >1,000 but <100,000. |
| ################################################################################ |
| - id: 2008 |
| metric_name: nand_read_attempt_failures |
| metric_type: OCCURRENCE |
| metric_semantics: [USAGE_COUNTING] |
| reports: |
| ############################################################################ |
| # A daily histogram of read failures per device. |
| ############################################################################ |
| - report_name: failures_histogram |
| id: 1 |
| report_type: UNIQUE_DEVICE_HISTOGRAMS |
| local_aggregation_period: WINDOW_1_DAY |
| int_buckets: |
| exponential: |
| # 0, 1, 2, 4, 8... 32768, 65536 |
| floor: 0 |
| num_buckets: 18 |
| initial_step: 1 |
| step_multiplier: 2 |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_selection: SELECT_LAST |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Number of failed reads where we exhausted all attempts |
| # |
| # A count of how many reads where no attempts sent to the nand succeed. |
| # |
| # For insight into how fleetwide failures might be clustered in a subset of |
| # devices, and how bad some of those devices might be. Expecting daily values |
| # to be normally >1 but <100. Could in theory be higher if instances of this |
| # didn't usually result in the system becoming inoperable. |
| ################################################################################ |
| - id: 2009 |
| metric_name: nand_read_attempts_exhausted |
| metric_type: OCCURRENCE |
| metric_semantics: [USAGE_COUNTING] |
| reports: |
| ############################################################################ |
| # A daily histogram of exhausted read attempts |
| ############################################################################ |
| - report_name: exhausted_histogram |
| id: 1 |
| report_type: UNIQUE_DEVICE_HISTOGRAMS |
| local_aggregation_period: WINDOW_1_DAY |
| int_buckets: |
| exponential: |
| # 0, 1, 2, 4, 8... 64, 128 |
| floor: 0 |
| num_buckets: 9 |
| initial_step: 1 |
| step_multiplier: 2 |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_selection: SELECT_LAST |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Number of bad blocks found on a device |
| # |
| # A count of how many blocks are found bad as marked by the initial factory |
| # mark, and how many are found and marked as bad during run time. |
| # |
| # For insight into the fleetwide state regarding the availability of the spare |
| # block pools, determined by how many blocks are currently marked bad. Also the |
| # split of initial bad blocks vs running bad blocks. For current use-cases 44 |
| # is really the maximum. |
| ################################################################################ |
| - id: 2010 |
| metric_name: ftl_bad_blocks |
| metric_type: INTEGER |
| metric_semantics: [USAGE_COUNTING] |
| metric_units_other: "blocks" |
| metric_dimensions: |
| - dimension: bad_block_type |
| event_codes: |
| 0: Unknown |
| 1: INITIAL |
| 2: RUNNING |
| reports: |
| ############################################################################ |
| # A daily histogram of bad block information per device. |
| ############################################################################ |
| - report_name: bad_blocks_histogram |
| id: 1 |
| report_type: UNIQUE_DEVICE_HISTOGRAMS |
| local_aggregation_period: WINDOW_1_DAY |
| local_aggregation_procedure: MAX_PROCEDURE |
| int_buckets: |
| exponential: |
| # 0, 1, 2, 4, 8... 32, 64 |
| floor: 0 |
| num_buckets: 8 |
| initial_step: 1 |
| step_multiplier: 2 |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_selection: SELECT_LAST |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Intermittent failures reading the last page in a map block |
| # |
| # Counts of occurrences where the FTL has encountered a failure reading a page |
| # in a map block that has subsequently succeeded. A dimension indicates the |
| # reason for the initial failure. |
| ################################################################################ |
| - id: 2011 |
| metric_name: ftl_map_block_end_page_failures |
| metric_type: OCCURRENCE |
| metric_semantics: [USAGE_COUNTING] |
| metric_dimensions: |
| - dimension: reason |
| event_codes: |
| # Whilst using event code 0 is not recommended, the reporting code always |
| # reports a code and reason 0 is used. This metric is also intended to |
| # be temporary. |
| 0: Reason 0 (invalid page) |
| 1: Reason 1 (erased page) |
| 2: Reason 2 (uncorrectable ecc error) |
| 3: Reason 3 (bad block-count) |
| 4: Reason 4 (bad map page number) |
| 5: Reason 5 (reserved for future use) |
| 6: Reason 6 (reserved for future use) |
| 7: Reason 7 (reserved for future use) |
| 8: Reason 8 (reserved for future use) |
| 9: Reason 9 (reserved for future use) |
| reports: |
| ############################################################################ |
| # A daily histogram of bad block information per device. |
| ############################################################################ |
| - report_name: counts |
| id: 1 |
| report_type: FLEETWIDE_OCCURRENCE_COUNTS |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| # Next ID: 2012 |
| |
| ################################################################################ |
| |
| # FVM Metrics |
| # Reserved Range: [3000, 4000) |
| |
| ################################################################################ |
| # FVM: Slice Allocation Counts |
| # |
| # Collects per-partition slice allocation counts from fvm. This assumes there are two fvm |
| # partitions - blobfs and minfs. |
| # |
| # Sampling Frequency: 1 sample per hour. |
| # |
| # |partition|: The partition allocating the slices. |
| ################################################################################ |
| - id: 3001 |
| metric_name: fvm_slice_allocations_migrated |
| metric_type: OCCURRENCE |
| metric_semantics: [USAGE_COUNTING] |
| metric_dimensions: |
| - dimension: partition |
| event_codes: |
| 0: Blobfs |
| 1: Minfs |
| reports: |
| - report_name: fvm_slice_allocations |
| id: 3000 |
| report_type: HOURLY_VALUE_NUMERIC_STATS |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| # Next ID: 3002 |
| |
| ################################################################################ |
| |
| # Space Metrics |
| # Reserved Range: [4000, 5000) |
| |
| ################################################################################ |
| # Data Bytes Used |
| # |
| # Used to track the number of bytes a partition is using for storing data. |
| # This is generally updated on a regular interval via Lapis, which queries |
| # filesystem information from fshost. |
| ################################################################################ |
| - id: 4001 |
| metric_name: data_bytes_used_migrated |
| metric_type: INTEGER |
| metric_units: BYTES |
| metric_semantics: [USAGE_COUNTING] |
| metric_dimensions: &space_dimensions |
| - *source_dimension |
| reports: |
| - report_name: per_device_daily_max |
| id: 1 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: MAX_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Data Bytes Allocated |
| # |
| # Used to track the number of bytes a partition has allocated (free + used) for |
| # storing data. |
| # This is generally updated on a regular interval via Lapis, which queries |
| # filesystem information from fshost. |
| ################################################################################ |
| - id: 4003 |
| metric_name: data_bytes_allocated_migrated |
| metric_type: INTEGER |
| metric_units: BYTES |
| metric_semantics: [USAGE_COUNTING] |
| metric_dimensions: *space_dimensions |
| reports: |
| - report_name: per_device_daily_max |
| id: 1 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: MAX_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Inodes Used |
| # |
| # Used to track the number of inodes a partition has in use. |
| # This is generally updated on a regular interval via Lapis, which queries |
| # filesystem information from fshost. |
| ################################################################################ |
| - id: 4005 |
| metric_name: inodes_used_migrated |
| metric_type: INTEGER |
| metric_units_other: "inodes" |
| metric_semantics: [USAGE_COUNTING] |
| metric_dimensions: *space_dimensions |
| reports: |
| - report_name: per_device_daily_max |
| id: 1 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: MAX_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Inodes Allocated |
| # |
| # Used to track the number of inodes a partition has capacity for (free + used). |
| # This is generally updated on a regular interval via Lapis, which queries |
| # filesystem information from fshost. |
| ################################################################################ |
| - id: 4007 |
| metric_name: inodes_allocated_migrated |
| metric_type: INTEGER |
| metric_units_other: "inodes" |
| metric_semantics: [USAGE_COUNTING] |
| metric_dimensions: *space_dimensions |
| reports: |
| - report_name: per_device_daily_max |
| id: 1 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: MAX_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| |
| ################################################################################ |
| # Out-Of-Space Events |
| # |
| # Per-filesystem count of times a device fails to extend the underlying volume |
| # after running out of space. |
| # |
| # Can only increase at most 1 every 5 minutes, so the daily max is 288. |
| ################################################################################ |
| - id: 4008 |
| metric_name: out_of_space_events |
| metric_type: OCCURRENCE |
| metric_semantics: [USAGE_COUNTING] |
| metric_dimensions: |
| - *source_dimension |
| reports: |
| - report_name: per_device_daily_counts |
| id: 1 |
| report_type: UNIQUE_DEVICE_HISTOGRAMS |
| system_profile_selection: SELECT_LAST |
| local_aggregation_period: WINDOW_1_DAY |
| int_buckets: |
| exponential: |
| # 0, 1, 2, 4, 8... 64, 128, 256 |
| floor: 0 |
| num_buckets: 10 |
| initial_step: 1 |
| step_multiplier: 2 |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| # Next ID: 4009 |
| |
| ################################################################################ |
| |
| # Fxfs Metrics |
| # Reserved Range: [8000, 9000) |
| |
| ################################################################################ |
| |
| ################################################################################ |
| # Bytes Reserved |
| # |
| # Number of bytes which are internally reserved by Fxfs. See Fxfs' |
| # object_manager.rs for details. |
| # |
| # Updated via Sampler which (at the time of writing) polls every 5 minutes. |
| ################################################################################ |
| - id: 8000 |
| metric_name: fxfs_bytes_reserved |
| metric_type: INTEGER |
| metric_units: BYTES |
| metric_semantics: [DATA_SIZE] |
| reports: |
| - report_name: per_device_daily_max |
| id: 1 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: MAX_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| # Same as above, but as a percentage of the total disk size (integer, 0-100) |
| - id: 8001 |
| metric_name: fxfs_bytes_reserved_percentage |
| metric_type: INTEGER |
| metric_units_other: "percent" |
| metric_semantics: [DATA_SIZE] |
| reports: |
| - report_name: per_device_daily_max |
| id: 1 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: MAX_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Bytes Borrowed |
| # |
| # Percentage of reserved bytes which are used for borrows. The inverse of this |
| # percentage gives the percentage which is used for metadata. See Fxfs' |
| # object_manager.rs for details. |
| # |
| # Updated via Sampler which (at the time of writing) polls every 5 minutes. |
| ################################################################################ |
| - id: 8002 |
| metric_name: fxfs_bytes_borrowed_to_reserved_percentage |
| metric_type: INTEGER |
| metric_units_other: "percent" |
| metric_semantics: [DATA_SIZE] |
| reports: |
| - report_name: per_device_daily_max |
| id: 1 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: MAX_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Journal Size |
| # |
| # Current size of the journal in bytes. (The journal may be slightly larger in |
| # practice, but the extra bytes are discardable at any time and will usually be |
| # trimmed on the next transaction). See Fxfs' journal.rs for details. |
| # |
| # Updated via Sampler which (at the time of writing) polls every 5 minutes. |
| ################################################################################ |
| - id: 8003 |
| metric_name: fxfs_journal_size |
| metric_type: INTEGER |
| metric_units: BYTES |
| metric_semantics: [DATA_SIZE] |
| reports: |
| - report_name: per_device_daily_max |
| id: 1 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: MAX_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| # Same as above, but as a percentage of the total disk size (integer, 0-100). |
| - id: 8004 |
| metric_name: fxfs_journal_size_percentage |
| metric_type: INTEGER |
| metric_units_other: "percent" |
| metric_semantics: [DATA_SIZE] |
| reports: |
| - report_name: per_device_daily_max |
| id: 1 |
| report_type: UNIQUE_DEVICE_NUMERIC_STATS |
| local_aggregation_procedure: MAX_PROCEDURE |
| local_aggregation_period: WINDOW_1_DAY |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| # Next ID: 8005 |
| |
| ################################################################################ |
| |
| # Temporary Metrics |
| # Reserved Range: [9000, 10000) |
| |
| ################################################################################ |
| # Data partition upgrade |
| # |
| # Used to track the progress of a one-time data partition upgrade. |
| # Each `stage` is triggered as a one-off event when that stage begins, which |
| # occurs during boot. Since the overall procedure is intended to be a one-time |
| # event, this should only trigger once per device. |
| ################################################################################ |
| - id: 9003 |
| metric_name: data_partition_upgrade_migrated |
| metric_type: OCCURRENCE |
| metric_semantics: [USAGE_COUNTING] |
| metric_dimensions: |
| - dimension: stage |
| event_codes: |
| 0: Unknown |
| 1: Skipped |
| 2: DetectedFailedUpgrade |
| 3: ReadOldData |
| 4: WriteNewData |
| 5: Done |
| max_event_code: 15 |
| reports: |
| - report_name: counts |
| id: 1 |
| report_type: FLEETWIDE_OCCURRENCE_COUNTS |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Recovered Out-Of-Space Events |
| # |
| # Indicates a successful volume extension after performing a corrective action. |
| # Used to track how many times a filesystem was successfully able to recover |
| # from a volume extension failure (e.g. by force-flushing the journal). |
| # |
| # Can only increase at most 1 every 5 minutes, so the daily max is 288. |
| ################################################################################ |
| - id: 9004 |
| metric_name: recovered_space_from_sync_events |
| metric_type: OCCURRENCE |
| metric_semantics: [USAGE_COUNTING] |
| metric_dimensions: |
| - *source_dimension |
| reports: |
| - report_name: per_device_daily_counts |
| id: 1 |
| report_type: UNIQUE_DEVICE_HISTOGRAMS |
| system_profile_selection: SELECT_LAST |
| local_aggregation_period: WINDOW_1_DAY |
| int_buckets: |
| exponential: |
| # 0, 1, 2, 4, 8... 64, 128, 256 |
| floor: 0 |
| num_buckets: 10 |
| initial_step: 1 |
| step_multiplier: 2 |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| ################################################################################ |
| # Minfs to Fxfs migration Events |
| # |
| # Indicates the status of any attempted disk-based migration flows at boot time. |
| # Used to track how successful the minfs to fxfs workflow is and give an |
| # indication of how often it has failed (and fallen back to minfs) in the field. |
| # |
| # This maps from ZX_ERR_* codes to events. |
| ################################################################################ |
| - id: 9005 |
| metric_name: minfs_to_fxfs_migration |
| metric_type: OCCURRENCE |
| metric_semantics: [USAGE_COUNTING] |
| metric_dimensions: |
| - dimension: err_code |
| event_codes: |
| 0: SUCCESS |
| 1: OUT_OF_SPACE |
| 2: OTHER_ERROR |
| reports: |
| ############################################################################ |
| # A daily histogram of migration failures per device. |
| ############################################################################ |
| - report_name: failures_histogram |
| id: 1 |
| report_type: UNIQUE_DEVICE_HISTOGRAMS |
| local_aggregation_period: WINDOW_1_DAY |
| int_buckets: |
| exponential: |
| # 0, 1, 2, 4, 8... 32768, 65536 |
| floor: 0 |
| num_buckets: 18 |
| initial_step: 1 |
| step_multiplier: 2 |
| privacy_level: NO_ADDED_PRIVACY |
| privacy_mechanism: DE_IDENTIFICATION |
| system_profile_selection: SELECT_LAST |
| system_profile_field: [BOARD_NAME, PRODUCT_NAME, SYSTEM_VERSION, CHANNEL] |
| meta_data: |
| max_release_stage: GA |
| expiration_date: "2025/04/05" |
| |
| # Next ID: 9006 |
| |
| # Deleted metric IDs that must not be reused. |
| deleted_metric_ids: [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,51,52,53,54,55,56,57,58,59,60,68,69,2000,2001,2002,3000,4000,4002,4004,4006,9000,9001,9002] |
