Manage statistics using the CLI
This page describes how to manage the statistics using the CLI.
Using the CLI, you can:
List statistics types
Command: weka stats list-types
Use the following command line to obtain statistics definition information:
weka stats list-types [<name-or-category>] [--show-internal]
Parameters
name-or-category
Limit the output to a specific statistic name, category, or category label (in parenthesis).
show-internal
Include internal statistics in the output.
False
Examples
Filter by category label (
"api statistics")
$ weka stats list-types "api statistics"
CATEGORY CATEGORY LABEL IDENTIFIER DESCRIPTION LABEL TYPE UNIT PARAMETERS RELATED RATE PARAMETER PERMISSION FOR NODE TYPE CAN ACCUMULATE HISTOGRAM HISTOGRAM UNIT
api api statistics TOTAL_2xx_RQ Total 2xx requests Total 2xx requests (total) Absolute Requests USER MANAGEMENT True False
api api statistics TOTAL_3xx_RQ Total 3xx requests Total 3xx requests (total) Absolute Requests USER MANAGEMENT True False
api api statistics TOTAL_429_RQ Total 429 requests Total 429 requests (total) Absolute Requests USER MANAGEMENT True False
api api statistics TOTAL_4xx_RQ Total 4xx requests Total 4xx requests (total) Absolute Requests USER MANAGEMENT True False
api api statistics TOTAL_5xx_RQ Total 5xx requests Total 5xx requests (total) Absolute Requests USER MANAGEMENT True FalseFilter by name and category (
api)
$ weka stats list-types api
CATEGORY CATEGORY LABEL IDENTIFIER DESCRIPTION LABEL TYPE UNIT PARAMETERS RELATED RATE PARAMETER PERMISSION FOR NODE TYPE CAN ACCUMULATE HISTOGRAM HISTOGRAM UNIT
api api statistics TOTAL_2xx_RQ Total 2xx requests Total 2xx requests (total) Absolute Requests USER MANAGEMENT True False
api api statistics TOTAL_3xx_RQ Total 3xx requests Total 3xx requests (total) Absolute Requests USER MANAGEMENT True False
api api statistics TOTAL_429_RQ Total 429 requests Total 429 requests (total) Absolute Requests USER MANAGEMENT True False
api api statistics TOTAL_4xx_RQ Total 4xx requests Total 4xx requests (total) Absolute Requests USER MANAGEMENT True False
api api statistics TOTAL_5xx_RQ Total 5xx requests Total 5xx requests (total) Absolute Requests USER MANAGEMENT True False
ops_s3 Operations (S3) API_FAILURES Total of failures per API Failures per API (total) Accumulated Ops api USER MANAGEMENT False False
ops_s3 Operations (S3) API_OPS Total of Ops per API Ops per API (total) Accumulated Ops api USER MANAGEMENT False False
ops_s3 Operations (S3) API_TTFB Time To First Byte per API TTFB per API (sum) Accumulated Milliseconds api USER MANAGEMENT False False
ops_s3 Operations (S3) API_TTLB Time To Last Byte per API TTLB per API (sum) Accumulated Milliseconds api USER MANAGEMENT False FalseInclude internal stats and filter by name and category (
config)
$ weka stats list-types config --show-internal
CATEGORY CATEGORY LABEL IDENTIFIER DESCRIPTION LABEL TYPE UNIT PARAMETERS RELATED RATE PARAMETER PERMISSION FOR NODE TYPE CAN ACCUMULATE HISTOGRAM HISTOGRAM UNIT
cluster Processes PEER_CONFIGURE_FAILURES How many times the node failed to configure peers to sync with them Peer configure failures detected (total) Absolute Peer configure failures WEKA ANY True False
config Config AVERAGE_CHANGES_IN_CHANGESET The average number of changes in a changeset Average (total) Rate Changes/Sec WEKA MANAGEMENT True False
config Config AVERAGE_CHANGES_IN_GENERATION The average number of changes in a generation Average (total) Rate Changes/Sec WEKA MANAGEMENT True False
config Config BACKEND_NODE_REJOIN_TIME The number of backends rejoin attempts per completion time range Backend node rejoin time (total) Histogram Number of rejoins WEKA MANAGEMENT True True mSecs
config Config CHANGESET_COMMIT_LATENCY The average latency of committing a configuration changeset Commit changeset latency (per changeset commits) Measured Microseconds TOTAL_CHANGESETS_COMMITTED WEKA MANAGEMENT True False
config Config CLIENT_NODE_REJOIN_TIME The number of clients rejoin attempts per completion time range Client node rejoin time (total) Histogram Number of rejoins WEKA MANAGEMENT True True mSecs
config Config CONFIG_PROPAGATION_LATENCY The latencies of propagation of a configuration generation Config Propagation latency (total) Histogram Generation WEKA MANAGEMENT True True Milliseconds
config Config FetchLocalStateChangesCallType_INTERNAL_CONTINUE Number of RPC calls to fetch local-state LRU changes, per overlay node type & call type Number of RPC calls to fetch local-state LRU changes per INTERNAL & CONTINUE (total) Absolute RPC Calls fetchLocalStateChangesCallType, overlayTreeNodeType WEKA MANAGEMENT True False
config Config FetchLocalStateChangesCallType_INTERNAL_RESTART_FROM_TAIL Number of RPC calls to fetch local-state LRU changes, per overlay node type & call type Number of RPC calls to fetch local-state LRU changes per INTERNAL & RESTART_FROM_TAIL (total) Absolute RPC Calls fetchLocalStateChangesCallType, overlayTreeNodeType WEKA MANAGEMENT True False
config Config FetchLocalStateChangesCallType_INTERNAL_RETRY_LAST_REQUEST Number of RPC calls to fetch local-state LRU changes, per overlay node type & call type Number of RPC calls to fetch local-state LRU changes per INTERNAL & RETRY_LAST_REQUEST (total) Absolute RPC Calls fetchLocalStateChangesCallType, overlayTreeNodeType WEKA MANAGEMENT True False
config Config FetchLocalStateChangesCallType_LEAF_CONTINUE Number of RPC calls to fetch local-state LRU changes, per overlay node type & call type Number of RPC calls to fetch local-state LRU changes per LEAF & CONTINUE (total) Absolute RPC Calls fetchLocalStateChangesCallType, overlayTreeNodeType WEKA MANAGEMENT True False
config Config FetchLocalStateChangesCallType_LEAF_RESTART_FROM_TAIL Number of RPC calls to fetch local-state LRU changes, per overlay node type & call type Number of RPC calls to fetch local-state LRU changes per LEAF & RESTART_FROM_TAIL (total) Absolute RPC Calls fetchLocalStateChangesCallType, overlayTreeNodeType WEKA MANAGEMENT True False
config Config FetchLocalStateChangesCallType_LEAF_RETRY_LAST_REQUEST Number of RPC calls to fetch local-state LRU changes, per overlay node type & call type Number of RPC calls to fetch local-state LRU changes per LEAF & RETRY_LAST_REQUEST (total) Absolute RPC Calls fetchLocalStateChangesCallType, overlayTreeNodeType WEKA MANAGEMENT True False
config Config GENERATION_COMMIT_LATENCY The average latency of committing a configuration generation to the RAFT log Commit changeset latency (per changeset commits) Measured Microseconds TOTAL_GENERATIONS_COMMITTED WEKA MANAGEMENT True False
config Config HEARTBEAT_PROCESSING_TIME The number of non-leader heartbeats per processing time range Non-leader heartbeat processing time (total) Histogram Number of heartbeats WEKA MANAGEMENT True True secs
config Config HEARTBEAT_PROCESSING_TIME_OLD The number of non-leader heartbeats per processing time range (OLD) Non-leader heartbeat processing time (total) Histogram Number of heartbeats WEKA MANAGEMENT True True secs
config Config HISTOGRAM_LEADER_ITERATION_WAIT_DURATION_CONFIG_ALIGNMENT Wait duration of leader iteration for all nodes to align on latest configuration generation Leader Iteration Wait for Configuration Propagation (total) Histogram Leader iteration wait time WEKA MANAGEMENT True True mSecs
config Config LEADER_HEARTBEAT_PROCESSING_TIME The number of leader heartbeats per processing time range Leader heartbeat processing time (total) Histogram Number of heartbeats WEKA MANAGEMENT True True secs
config Config LEADER_HEARTBEAT_PROCESSING_TIME_OLD The number of leader heartbeats per processing time range (OLD) Leader heartbeat processing time (total) Histogram Number of heartbeats WEKA MANAGEMENT True True secs
config Config LOCALSTATE_AGGREGATION_LATENCY This period between clockSkewReportTime table's update by a management process and the time the leader sees it LocalState aggregation latency (total) Histogram Time is taken to aggregate LocalState in milliseconds WEKA MANAGEMENT True True msecs
config Config LOCAL_STATS_FETCH_GENERATION_LAGGING The number of local-state generations that the parent fetch request still needs to read local-state generation dirty LRU length (average) Accumulated local-state generations overlayParentIndex WEKA MANAGEMENT False False
config Config OVERLAY_FULL_SHIFTS The number of entire overlay shifts Full overlay shifts (total) Absolute Changes WEKA MANAGEMENT True False
config Config OVERLAY_INCREMENTAL_SHIFTS The number of incremental overlay shifts Incremental overlay shifts (total) Absolute Changes WEKA MANAGEMENT True False
config Config OVERLAY_TRACKER_INCREMENTALS The number of incremental OverlayTracker applications Incremental OverlayTracker applications (total) Absolute Changes WEKA MANAGEMENT True False
config Config OVERLAY_TRACKER_RESYNCS The number of OverlayTracker full-resyncs OverlayTracker resyncs (total) Absolute Changes WEKA MANAGEMENT True False
config Config TOTAL_CHANGESETS_COMMITTED The total number of committed changesets Changeset commits (total) Absolute Change Sets WEKA MANAGEMENT True False
config Config TOTAL_COMMITTED_CHANGES The total number of committed configuration change sets Changes (total) Absolute Changes WEKA MANAGEMENT True False
config Config TOTAL_CONFIG_SNAPSHOT_PULLS The total number of config snapshot pulls Config snapshot pulls (total) Absolute Pulls WEKA MANAGEMENT True False
config Config TOTAL_GENERATIONS_COMMITTED The number of committed generations Config generation commits (total) Absolute Generations WEKA MANAGEMENT True False
raft RAFT Configuration_LEADER_CHANGES Changes of leader Configuration Leader Changes (total) Absolute Changes councilType WEKA ANY True False
raft RAFT Configuration_REQUESTS_COMPLETED Requests to leader completed successfully Configuration Requests Completed (total) Absolute Requests councilType WEKA ANY True FalseView statistics in real-time
Command: weka stats realtime
Use the following command line to obtain the current performance-related statistics of the processes in a one-second interval:
weka stats realtime [<process-ids>] [--raw-units] [--UTC]
Parameters
process-ids
Only show real-time stats of the specified processes in a comma-separated list.
raw-units
Print values in raw units such as bytes and seconds.
Readable format. Examples: 1KiB 234MiB 2GiB.
UTC
Print times in UTC.
Server's local time.
View statistics over time
Command: weka stats
The collected statistics can help analyze system performance and determine the source of issues that may occur during WEKA system runs. Statistics are divided according to categories. When selecting a category, a list of the possible statistics is displayed, from which you can select the specific statistics.
With the exception of real-time statistics, WEKA averages all statistics over one-minute intervals. Therefore, the total value or other aggregates relate to a specific minute.
Use the following command line to manage filters and read statistics:
weka stats [--start-time <start-time>] [--end-time <end-time>] [--interval interval] [--resolution-secs resolution-secs] [--category category][--stat stat] [--process-ids process-ids] [--param param] [--accumulated] [--per-process] [--no-zeros] [--show-internal] [--raw-units] [--UTC]
Parameters
start-time
Start time of the reported period.
Format examples: 5m, -5m, -1d, -1w, 1:00, 01:00, 18:30, 18:30:07, 2018-12-31 10:00, 2018/12/31 10:00, 2018-12-31T10:00, 9:15Z, 10:00+2:00.
-1m
end-time
End time of the reported period.
Format examples: 5m, -5m, -1d, -1w, 1:00, 01:00, 18:30, 18:30:07, 2018-12-31 10:00, 2018/12/31 10:00, 2018-12-31T10:00, 9:15Z, 10:00+2:00.
Current time
interval*
Period of time to be reported. Valid interval in seconds (positive integer).
resolution-secs
Length of each interval in the reported period. The value must be a multiple of 60 seconds
60
category
A specific category for retrieving statistics. Valid categories: CPU, Object Store, Operations, Operations (NFS), Operations (Driver), SSD.
All
stat
The names of the statistics to retrieve.
All
process-ids
A valid process ID.
All
param
For parameterized statistics, it retrieves only the instantiations where the specified parameter has the specified value.
Format: key:val
Example for multiple values:
'--param method:putBlocks --param method:initBlock'
accumulated
Display accumulated statistics instead of rate statistics.
False
per-process
Do not aggregate statistics across processes.
False
no-zeros
Filters out results where the value is 0.
False
show-internal
Display internal statistics.
False
raw-units
Print values in raw units, such as bytes and seconds.
Readable format. (for example: 1KiB 234MiB 2GiB)
UTC
Print times in UTC.
Server's local time.
Set statistics retention
Command: weka stats retention set
Use the following command line to set the statistics retention period.
weka stats retention set <--days days> [--dry-run]
Parameters
days*
Number of days to keep the statistics. Ensure sufficient free disk space per server and the specified number of days.
dry-run
Only tests the required capacity per the retention period.
Use weka stats retention status to view the current retention and weka stats retention restore-default to restore the default retention settings.
Last updated