Enabling JMX counters for performance monitoring

Bitbucket Data Center と Server の管理

このページの内容

お困りですか?

アトラシアン コミュニティをご利用ください。

コミュニティに質問

This article describes how to expose JMX MBeans within Bitbucket Data Center and Server for monitoring with a JMX client.

Related reading: Understanding JMX (Oracle)

On this page:

JMX とは何か

JMX (Java Management eXtensions) is a technology for monitoring and managing Java applications. JMX uses objects called MBeans (Managed Beans) to expose data and resources from your application.

Why would I want to enable JMX monitoring within Bitbucket?

For large Bitbucket instances, enabling JMX allows you to more easily monitor the consumption of application resources. This enables you to make better decisions about how to maintain and optimize machine resources.

How to enable JMX monitoring?

Starting from Bitbucket 8.0, Git operations are run in the Bitbucket Mesh sidecar by default. That’s why, to enable full JMX monitoring for Bitbucket 8.0 and later, you should enable JMX for both the Bitbucket main process and the Mesh sidecar process.

To enable JMX for the Bitbucket main process:

  1. Go to Administration > General configuration > Monitoring.

  2. Use the JMX monitoring toggle to enable or disable the feature.

Learn how to enable JMX monitoring on the Mesh sidecar

JMX を使用して何を監視できますか?

JMX counters allow you to monitor various statistics across Bitbucket. Select the one that interests you and find out more about it in the document.

Mail performance

Mail statistics (com.atlassian.bitbucket:name=MailStatistics)

名前説明
AverageMessageSizeAverage size (in bytes) of messages sent
LargestMessageSentLargest message that has been sent (in bytes)
LastMessageFailureDate of the last failure to send a message
LastMessageSuccessDate of the last successful message send operation
LastQueueFullEventLast time the message queue was full
QueueFullEventCountNumber of times the message queue was full
QueueUsageQueue usage as a fraction, 0.0d indicates empty and 1.0d indicates full
QueuedMessagesCountCurrent count of queued (unsent) messages
QueuedMessagesSizeCurrent size (in bytes) of the queued (unsent) messages
TotalMailDataSentTotal size (in bytes) of messages sent
TotalMessagesFailedTotal number of messages that failed to send
TotalMessagesSentTotal number of messages sent

診断

Alerts total (com.atlassian.diagnostics:type=Alerts,name=Total)

名前説明
LatestAlertTimestampTimestamp of the most recent alert
TotalCountTotal number of alerts since the JVM was started
ErrorCountNumber of alerts of severity ERROR since the JVM was started
InfoCountNumber of alerts of severity INFO since the JVM was started
WarningCountNumber of alerts of severity WARNING since the JVM was started

Plugin (com.atlassian.diagnostics:type=Alerts,Category=Plugin,name=${PLUGIN_NAME})

名前説明
LatestAlertTimestampTimestamp of the most recent alert
TotalCountTotal number of alerts since the JVM was started
ErrorCountNumber of alerts of severity ERROR since the JVM was started
InfoCountNumber of alerts of severity INFO since the JVM was started
WarningCountNumber of alerts of severity WARNING since the JVM was started
PluginNamePlugin name, if available

Issue (com.atlassian.diagnostics:type=Alerts,Category=Issue,name=${ISSUE_ID})

名前説明
LatestAlertTimestampTimestamp of the most recent alert
コンポーネントComponent the issue is defined for
countNumber of alerts for the issue since the JVM was started
深刻度Issue's severity
説明 Issue's description

Hosting statistics

プロトコルObject name
sshcom.atlassian.bitbucket:name=SshHostingStatistics
httpcom.atlassian.bitbucket:name=HttpHostingStatistics

Hosting statistic attributes

All the hosting statistics attributes are monotonically increasing since the JVM was restarted

名前説明
CloneCacheBypassClone requests that have bypassed the scm-cache
CloneCacheHitClone requests served from the scm-cache
CloneCacheMissClone requests that could not be served from the scm-cache
CloneReadBytes read from clients during clone operations
CloneRequestCountNumber of clone requests served
CloneWrittenbytes written to clients during <i>clone</i> operations
FetchReadBytes read from clients during <i>fetch</i> operations
FetchRequestCountNumber of fetch requests served
FetchWrittenBytes written to clients during fetch operations
RequestsTotal number of requests served
TotalBytesReadTotal bytes read from clients
TotalBytesWrittenTotal bytes written to clients

Webhooks statistics (com.atlassian.webhooks:name=Webhooks)

名前説明
PublishCountA count of the total number of events that could trigger webhooks
(A publish may create many dispatches)
DispatchSuccessCountTotal number of webhooks to fire successfully with a successful HTTP response
DispatchRejectedCountA count of the number of webhook dispatches that were rejected for execution
DispatchLastRejectedTimestampThe last time a webhook was rejected, either from circuit breaking, or due to too many webhooks being in flight
DispatchInFlightCountTotal number of dispatches that have been triggered and are awaiting resolution
DispatchFailureCountTotal number of webhooks that fired successfully, but the HTTP response indicates a failure
(non 2xx code)
DispatchErrorCountTotal number of webhooks to have had an error while they were being dispatched
DispatchCountTotal number of webhooks to have been dispatched

Thread pools

Thread pool説明 Object name
BuildActionsThreadPoolThreads that handles Integrated CI/CD build actionscom.atlassian.bitbucket.thread-pools:name=BuildActionsThreadPool
EventThreadPoolThreads that dispatch events to @EventListenermethodscom.atlassian.bitbucket.thread-pools:name=EventThreadPool
IoPumpThreadPoolThreads that handle blocking process I/Ocom.atlassian.bitbucket.thread-pools:name=IoPumpThreadPool
NioPumpThreadPoolThreads that handle nonblocking process I/Ocom.atlassian.bitbucket.thread-pools:name=NioPumpThreadPool
ScheduledThreadPoolThread pool that takes care of several miscellaneous scheduled taskscom.atlassian.bitbucket.thread-pools:name=ScheduledThreadPool
QueuedThe number of requests currently waiting for an available ticketN/A

Thread pool attributes

名前

説明

ActiveCountReturns the approximate number of threads that are actively executing tasks
MaximumPoolSizeReturns the maximum allowed number of threads
PoolSizeReturns the current number of threads in the pool
QueueLengthThe number of tasks awaiting execution by the thread pool
LargestPoolSizeThe largest number of threads that have ever been simultaneously in the pool
CompletedTaskCountThe approximate total number of tasks that have completed execution. Because the states of tasks and threads may change dynamically during computation, the returned value is only an approximation, but one that does not ever decrease across successive calls

Repositories (com.atlassian.bitbucket:name=Repositories)

名前説明
countNumber of repositories currently configured across all projects

Scm Statistics (com.atlassian.bitbucket:name=ScmStatistics)

名前説明
PullsNumber of scm pulls serviced by this instance since it was started
PushesNumber of scm pushes received by this instance is it was started

Ticket statistics

Starting from Bitbucket 8.0, the Ticket metrics are absent from the com.atlassian.bitbucket property and must be fetched from the Bitbucket Mesh sidecar.

Learn how to enable JMX monitoring for the Mesh sidecar in BSERV-13623 - Getting issue details... STATUS

Bitbucket uses tickets as a mechanism for creating back pressure to prevent the system from being overloaded with requests. There are two types of tickets: hosting tickets and command tickets.

  • Hosting tickets (com.atlassian.bitbucket:name=HostingTickets) limits the number of SCM hosting operations, meaning pushes and pulls over HTTP or SSH, which may be running concurrently. 
  • Command tickets (com.atlassian.bitbucket:name=CommandTickets) limits the number of SCM commands, such as: git diff, git blame, or git rev-list, which may be running concurrently.

Bitbucket supports the following metrics for each ticket type.

名前説明
AvailableThe number of tickets available for acquisition (lower number means higher load)
名前The name of the ticket bucket either 'scm-command' or 'scm-hosting'
合計The maximum number of tickets that can be acquired concurrently before back-pressure is applied
UsedThe number of tickets that have been acquired (higher number means higher load)
LastRejection

The timestamp of the last rejected ticket, or null if no tickets have been rejected.

If this metric is missing, follow the public issue BSERV-19088 to stay updated on the fix.

OldestQueuedRequest

The timestamp at which the oldest queued request started waiting, or null if there are no queued requests.

If this metric is missing, follow the public issue BSERV-19088 to stay updated on the fix.

Event statistics (com.atlassian.bitbucket:name=EventStatistics)

名前説明
DispatchedCountTotal number of listener callbacks that have been performed. An event that is delivered to 10 listeners counts as 10 dispatches
LastRejectionDate of the last event being rejected, or null if no event has been rejected
PublishedCountTotal number of events delivered. An event that is delivered to 10 listeners counts as 1 event
QueueCapacityMaximum number of event callbacks that can be queued before events are rejected
QueueLengthNumber of event callbacks that have been queued but haven't been dispatched yet
RejectedCountTotal number of events that were not dispatched because the event queue was full
RemainingQueueCapacityRemaining number of event callbacks that can be queued before events are rejected

Cluster lock statistics (com.atlassian.bitbucket:name=ClusterLocks)

名前説明

LockedCount

Number of cluster locks that are currently held by this node

QueuedThreadCount

Number of threads on this node that are currently blocked waiting for a lock

TotalAcquiredCount

Total number of times a cluster lock was acquired on this node since startup

TotalAcquireErrorCount

Number of times an exception was thrown while trying to acquire a cluster lock on this node since startup

TotalAcquireTimeMillis

Total time in milliseconds that any thread on this node has spent acquiring a lock (including time blocked waiting for a lock to become available)

TotalReleasedCount

Total number of times a cluster lock was released on this node since startup

TotalReleaseErrorCount

Total number of times an exception was thrown while releasing a cluster lock on this node since startup

SSH session statistics (com.atlassian.bitbucket:name=SshSessions)

名前説明
ActiveSessionCountNumber of currently active SSH session
MaxActiveSessionCountHighest number of concurrently active SSH sessions since the last startup
SessionClosedCountTotal number of SSH sessions that have been closed since the last startup
SessionCreatedCountTotal number of SSH sessions that have been created since the last startup
SessionExceptionCountTotal number of SSH sessions that have been terminated because an exception was thrown from the SSH command run
SessionDisconnectedLimitCountTotal number of SSH sessions that have been disconnected due to maximum concurrent requests
SessionDisconnectedConnectionLostCountTotal number of SSH sessions that have been disconnected due to connection loss

Rate limiting statistics (com.atlassian.bitbucket:name=RateLimitStatistics)

名前説明
RejectedRequestCountThe number of rate limited requests
UserMapSizeThe number of token buckets currently in memory

Licensed users statistics

名前説明
LicenseCapacityThe maximum number of users for the existing license
LicensedUserCountThe number of licensed users for the existing license

Third-party library attributes

Bitbucket exposes the JMX attributes from number of third party libraries. Listed below is a sample of the attributes that are particularly interesting from an operations perspective. 

HikariCP (com.zaxxer.hikari:type=Pool (bitbucket))

名前

説明

ActiveConnections

Active Connections (in use)

IdleConnections

Idle Connection count

ThreadsAwaitingConnection

The number of threads waiting for a connection (when all available connections are in use)

TotalConnectionsTotal Connections

Hibernate (org.hibernate.core:sessionFactory=bitbucket.core,serviceRole=org.hibernate.stat.Statistics,serviceType=org.hibernate.stat.internal.ConcurrentStatisticsImpl)

名前説明

QueryCacheHitCount

Global number of cached queries successfully retrieved from cache

QueryCacheMissCount

Global number of cached queries not found in cache

SecondLevelCacheHitCount

Global number of cacheable entities/collections successfully retrieved from the cache

SecondLevelCacheMissCount

Global number of cacheable entities/collections not found in the cache and loaded from the database

Expose JMX MBeans within Bitbucket

To enable Bitbucket to publish specific statistics using JMX:

  1. Locate and open the bitbucket.properties file in the <Bitbucket home directory>/shared directory. 
    1. Add this property to the file.

      jmx.enabled=true
    2. ファイルを保存して閉じます。
       
  2. Create a JMX password file for secure access to JMX monitoring.
     
  3. Modify the set-jmx-opts.sh file to enable Bitbucket to expose JMX Mbeans.

These changes will not take effect until Bitbucket is restarted.

Set up the JMX password file

Set up a JMX password file to secure access to JMX monitoring.

  1. Create a file named jmx.access.

    This file will contain password information. Ensure the file is only readable by the secure user Bitbucket will run under. However, note that if the user cannot read the file Bitbucket will fail to start.

  2. Edit the jmx.access file to include this property and save the file.

    monitorRole <password>

    If you wish to use a username other than monitorRole or controlRole you will need to modify the jmxremote.access file located in the /lib/management/ directory of the installed Java.

  3. Change ownership of jmx.access file,

    chown bitbucket:bitbucket <path>/jmx.access

    where bitbucket is the user that runs Bitbucket service.

  4. Change file permissions of jmx.access file.

    chmod 600 <path>/jmx.access

Modify the Bitbucket environment file

Modify the set-jmx-opts.sh (for Windows set-jmx-opts.bat) files to enable JMX monitoring:

  1. Within the bin directory, locate the file set-jmx-opts.sh (for Windows set-jmx-opts.bat) and change these properties.

    JMX_REMOTE_AUTH=password
    JMX_REMOTE_PORT=3333
    RMI_SERVER_HOSTNAME=-Djava.rmi.server.hostname=<hostname>
    JMX_PASSWORD_FILE=<path>/jmx.access
  2. Restart Bitbucket.

Docker

For Docker deployments, the properties can be passed as Docker environment variables:

Sample:

docker run -v /data/bitbucket:/var/atlassian/application-data/bitbucket \
  --name="bitbucket" \
  -d -p 7990:7990 -p 7999:7999 -p 3333:3333 \
  -e JMX_ENABLED=true \
  -e JMX_REMOTE_AUTH=password \
  -e JMX_REMOTE_PORT=3333 \
  -e JMX_REMOTE_RMI_PORT=3333 \
  -e RMI_SERVER_HOSTNAME=<hostname> \
  -e JMX_PASSWORD_FILE=<path>/jmx.access \
  atlassian/bitbucket

Verify JMX is configured correctly

These steps use JConsole to test that JMX has been configured correctly. JConsole is a utility that ships with the Oracle JDK.

  1. To start the jconsole utility, from a command line prompt enter

    jconsole

  2. Create a new JConsole connection with similar connection settings.

    bitbucketthe hostname of the instance to monitor

    3333

    the JMX port number previously configured.

    username, passwordvalues configured within the JMX password file jmx.access.
  3. 接続 をクリックします。

When configured correctly, you will see the following properties.

com.atlassian.bitbucket

Starting from Bitbucket 8.0, the Ticket metrics are absent from the com.atlassian.bitbucket property and must be fetched from the Bitbucket Mesh sidecar.

Learn how to enable JMX monitoring for the Mesh sidecar in BSERV-13623 - Getting issue details... STATUS

  • CommandTickets
  • HostingTickets
  • プロジェクト
  • リポジトリ
  • ScmStatistics
  • Tickets
  • EventStatistics
  • ClusterLocks

  • SshSessions

com.atlassian.bitbucket.thread-pools
  • EventThreadPool
  • IoPumpThreadPool
  • ScheduledThreadPool

Example performance dashboard

This dashboard was generated using Java Mission Control that ships with the Oracle JDK (since 1.7u40). See the documentation that comes with your JMX client of choice for more information. 

Configuring JMX to use SSL

You can find information about the options for configuring JMX to use SSL in the set-jmx-opts files. Comprehensive documentation is available from Oracle.

製品内診断メトリック

次のセクションを展開して、製品内診断で利用できるメトリックの詳細を学習してください。

To use these metrics, make sure you've first enabled JMX.

SSH connection metrics
Bean オブジェクト名メトリックの説明
com.atlassian.bitbucket:type=metrics,
category00=ssh,category01=sessions,
category02=active,name=value
  • ssh.sessions.active.value

The latest measure of the number of active SSH sessions.

com.atlassian.bitbucket:type=metrics,
category00=ssh,category01=sessions,
category02=active,name=statistics
  • ssh.sessions.active.statistics

Aggregated statistics of the number of active SSH sessions.

com.atlassian.bitbucket:type=metrics,
category00=ssh,category01=sessions,
category02=closed,name=value
  • ssh.sessions.closed.value

The latest measure of the number of closed SSH sessions.

com.atlassian.bitbucket:type=metrics,
category00=ssh,category01=sessions,
category02=created,name=value
  • ssh.sessions.created.value

The latest measure of the number of created SSH sessions.

com.atlassian.bitbucket:type=metrics,
category00=ssh,category01=sessions,
category02=created,name=statistics
  • ssh.sessions.created.statistics

Aggregated statistics of the number of created SSH sessions.

com.atlassian.bitbucket:type=metrics,
category00=ssh,category01=sessions,
category02=disconnected,category03=connection,
category04=lost,name=value
  • ssh.sessions.disconnected.connection.lost.value

The number of disconnected SSH requests due to too many concurrent connections.

com.atlassian.bitbucket:type=metrics,
category00=ssh,category01=sessions,
category02=disconnected,category03=connection,
category04=lost,name=statistics
  • ssh.sessions.disconnected.connection.lost.statistics

Aggregated statistics of the number of disconnected SSH requests due to too many concurrent connections.

com.atlassian.bitbucket:type=metrics,
category00=ssh,category01=sessions,
category02=disconnected,category03=limit,
name=value
  • ssh.sessions.disconnected.limit.value

The number of disconnected SSH requests due to connection lost.

com.atlassian.bitbucket:type=metrics,
category00=ssh,category01=sessions,
category02=disconnected,category03=limit,
name=statistics
  • ssh.sessions.disconnected.limit.statistics

Aggregated statistics of the number of disconnected SSH requests due to too many concurrent connections

com.atlassian.bitbucket:type=metrics,
category00=ssh,category01=sessions,
category02=error,name=counter
  • ssh.sessions.error.counter

The count of SSH connection failures since the last restart.

Ticket metrics
Bean オブジェクト名メトリックの説明
com.atlassian.bitbucket:type=metrics,
category00=tickets,category01=command,
category02=available,name=value
  • tickets.command.available.value

The latest measure of the number of available Command tickets.

com.atlassian.bitbucket:type=metrics,
category00=tickets,category01=command,
category02=available,name=statistics
  • tickets.command.available.statistics

Aggregated statistics of the number of available Command tickets.

com.atlassian.bitbucket:type=metrics,
category00=tickets,category01=command,
category02=used,name=value
  • tickets.command.used.value

The latest measure of the number of used Command tickets.

com.atlassian.bitbucket:type=metrics,
category00=tickets,category01=command,
category02=used,name=statistics
  • tickets.command.used.statistics

Aggregated statistics of the number of used Command tickets.

com.atlassian.bitbucket:type=metrics,
category00=tickets,category01=hosting,
category02=available,name=value
  • tickets.hosting.available.value

The latest measure of the number of available Hosting tickets.

com.atlassian.bitbucket:type=metrics,
category00=tickets,category01=hosting,
category02=available,name=statistics
  • tickets.hosting.available.statistics

Aggregated statistics of the number of available Hosting tickets.

com.atlassian.bitbucket:type=metrics,
category00=tickets,category01=hosting,
category02=used,name=value
  • tickets.hosting.used.value

The latest measure of the number of used Hosting tickets.

com.atlassian.bitbucket:type=metrics,
category00=tickets,category01=hosting,
category02=used,name=statistics
  • tickets.hosting.used.statistics

Aggregated statistics of the number of used Hosting tickets.

Storage metrics
Bean オブジェクト名メトリックの説明

com.atlassian.bitbucket:type=metrics,

category00=storage,category01=space,

category02=local,category03=free,name=value

  • storage.space.local.free.value

The value of the free space on the disk where the BB local storage is located.

com.atlassian.bitbucket:type=metrics,

category00=storage,category01=space,

category02=local,category03=used,name=value

  • storage.space.local.used.value

The value of the used space on the disk where the BB local storage is located.

com.atlassian.bitbucket:type=metrics,

category00=storage,category01=space,category02=local,

category03=caches,name=value

  • storage.space.local.caches.value

The size of the BITBUCKET_HOME/caches directory in bytes.

com.atlassian.bitbucket:type=metrics,

category00=storage,category01=space,

category02=local,category03=logs,name=value

  • storage.space.local.logs.value

The size of the BITBUCKET_HOME/log directory in bytes.

com.atlassian.bitbucket:type=metrics,

category00=storage,category01=space,category02=local,

category03=tmp,name=value

  • storage.space.local.tmp.value

The size of the BITBUCKET_HOME/tmp directory in bytes.

製品内診断モニタリングを有効にする

In-product monitoring is enabled by default. 

JMX ログ記録のポーリング間隔は 60 秒に設定されており、変更できません。

To manage it:

  1. Go to Administration > General configuration > Monitoring.
  2. Use the Enable in-product diagnostics toggle to enable or disable in-product monitoring.

ログ形式

Writing to atlassian-bitbucket-ipd-monitoring.log is done via log4j. Its configuration is managed in logback-spring.xml.

ログの内容

By default, a concise set of data is included in each log entry. An extended set of data can be logged by enabling the bitbucket.in.product.diagnostics.extended.logging dark feature flag.

Learn how to enable the dark feature flag

次の各表で、簡潔ログ形式と拡張ログ形式の構造をご確認ください。

The metrics in JMX always go in the extended data format

簡潔データ

MBean タイプProperties属性
カウンター

timestamp

ラベル

属性

_count
_value
統計

_99thPercentile

_max

_min

_mean

簡潔ログの行形式の例を表示
2023-01-13 11:51:13,106 IPDMONITORING {"timestamp":"1673610673","label":"DB.CONNECTION.POOL.NUMACTIVE.STATISTICS","attributes":{"_max":"2.0","_mean":"1.2436699769063984","_99thPercentile":"2.0","_count":"5","_min":"1.0"}}

拡張データ

The metrics in JMX always go in the extended data format. Learn more about the metric attributes

MBean タイプProperties属性
カウンター

timestamp

ラベル

属性

オブジェクト名

_count

_fifteenMinuteRate

_fiveMinuteRate

_meanRate

_oneMinuteRate

_rateUnit

_value

_number

統計

_50thPercentile

_75thPercentile

_95thPercentile

_98thPercentile

_99thPercentile

_999thPercentile

_count

_min

_max

_mean

_stdDev

_durationUnit

_fifteenMinuteRate

_fiveMinuteRate

_meanRate

_oneMinuteRate

_rateUnit

簡潔ログの行形式の例を表示
2022-09-06 18:38:48,015 IPDMONITORING {"timestamp":"1662453528","label":
"DB.CONNECTION.LATENCY.STATISTICS","objectName":
"com.atlassian.confluence:category00\u003ddb,category01\u003dconnection,category02\
u003dlatency,name\u003dstatistics,type\u003dmetrics",
"attributes":{"_oneMinuteRate":"0.02012497818617073","_50thPercentile":"0.0",
"_mean":"1.9379304604014412E-25","_max":"1.0","_stdDev":"4.40219315841711E-13",
"_98thPercentile":"0.0","_meanRate":"0.003612560785169162","_rateUnit":
"events/second","_99thPercentile":"0.0","_count":"16","_durationUnit":
"milliseconds","_75thPercentile":"0.0","_fiveMinuteRate":
"0.005912972095043379","_fifteenMinuteRate":"0.0037696657500141968",
"_999thPercentile":"0.0","_95thPercentile":"0.0","_min":"0.0"}}

指標属性の定義

指標属性の詳細は、次のセクションを展開してご確認ください。

回数の指標属性

属性

定義

_count

現在の時間枠で指標が出現した回数

_fifteenMinuteRate

直近 15 分間で指標が出現した回数

_fiveMinuteRate

直近 5 分間で指標が出現した回数

_meanRate

メーターの作成後にイベントが発生した平均レート

_oneMinuteRate

直近 1 分間で指標が出現した回数

_rateUnit

レートに使用される測定単位

次の属性にご注意ください: _oneMinuteRate_fiveMinuteRate_fifteenMinuteRate

_count は、測定値が時間の経過とともにどのように変化したかを示すものではありません。最新性は分単位のレートで提供されます。

値の指標属性

属性

定義

_value

指標の直近のサンプル値

_number

_value 属性と同じ値が含まれています

統計の指標属性

MBean タイプの統計の指標は、集計値とも呼ばれます。この指標は、一定期間で変更が行われた項目の統計を提供します。たとえば、メール キューで処理された項目や、エラー メール キューに追加された項目などです。

時間枠

特に明記されていない限り、集計値はスライディング時間枠で算出されます。これは、直近の 5 分間を概ねカバーします。

パーセンタイル値は、蓄積サンプリング技法で算出されます。この技法では、データ ストリーム全体を統計的に表す、小さくて管理しやすい値のセットを使用するため、メモリに保存する必要のあるデータの量が軽減します。

リセット

スライディング時間枠の外では、集計値が次のタイミングですべてリセットされます。

  • システムが再起動するたび。

  • JMX モニタリングまたは製品内診断指標が有効化されるたび。

その他の Data Center 製品の JMX モニタリングと製品内診断に関する詳細をご確認ください。

次の表に、統計の指標属性の定義を示します。

属性

定義

_50thPercentile

現在の時間枠における、すべての測定値の 50 % を下回る測定値。中央とも呼ばれます。

この属性は、中間の測定値を表す手段として、平均値の代わりになります。中央値は、平均値よりも外れ値によって歪む可能性が低くなります。

_75thPercentile

現在の時間枠内における、すべての測定値の 75 % を下回る測定値。第 3 四分位数

_95thPercentile

現在の時間枠内における、すべての測定値の 95 % を下回る測定値

_98thPercentile

現在の時間枠内における、すべての測定値の 98 % を下回る測定値

_99thPercentile

現在の時間枠内における、すべての測定値の 99 % を下回る測定値

_999thPercentile

現在の時間枠内における、すべての測定値の 999 % を下回る測定値

_count

現在の時間枠で指標が出現した回数

_min

現在の時間枠内における最小測定値

_max

現在の時間枠内における最大測定値_max_min の間の統計的範囲で値の変動性を測定する尺度

_mean

現在の時間枠における平均値。

この属性は、大きな外れ値の測定値によって歪む可能性があります。このような場合は、_50thPercentile が中間値としてより適切な指標を提供します。

_stdDev

データのばらつきを示す指標。

標準偏差が低い場合は、値がセットの平均値に近い傾向があることを示します。一方、標準偏差が高い場合は、値がより広い範囲に分散していることを示します。

_durationUnit

期間に使用される測定単位

_fifteenMinuteRate

直近 15 分間で指標が出現した回数

_fiveMinuteRate

直近 5 分間で指標が出現した回数

_meanRate

メーターの作成後にイベントが発生した平均レート

_oneMinuteRate

直近 1 分間で指標が出現した回数

_rateUnit

レートに使用される測定単位

次の属性にご注意ください: _oneMinuteRate_fiveMinuteRate_fifteenMinuteRate

_count は、測定値が時間の経過とともにどのように変化したかを示すものではありません。最新性は分単位のレートで提供されます。

最終更新日: 2024 年 2 月 7 日

この内容はお役に立ちましたか?

はい
いいえ
この記事についてのフィードバックを送信する
Powered by Confluence and Scroll Viewport.