立即与支持人员聊天
与支持团队交流

Foglight for DB2 (Cartridge) 7.1.0 - Getting Started Guide

Introduction to Foglight for DB2 LUW Using Reference
Dashboards and Views Collections and Metrics
CF Locks Collection CF Memory Pools Collection CF Resource Usage Collection Current Agents Details Collection Agents Summary Collection Current Locks Wait Collection Database Bufferpools Summary PS Collection Database Configuration Collection Database Tablespaces Collection Database Tablespaces Summary Collection Database Usability Collection Database Parameters Collection Database Instance Parameters Collection Database Instance Registry Variables Collection Database Tablespaces BP Collection FCM Activity Collection FS DB Storage Collection FS DB Storage Summary Collection FS Instance Storage Collection FS Storage Collection FS Storage Summary Collection General Activity Collection HADR Activity Collection Host Properties Collection OS General Collection Instance Activity Collection Instance General Collection Instance Nodes Collection Instance Usability Collection IO Activity Collection License Details Collection Log Activity Collection Log File Collection Log Message Collection Messages Count Collection Monitor Switches Collection OS DB2 General Process Collection Partition Tablespaces Summary Collection Partition Usability Collection pureScale Alerts Collection Top SQLs Collection Query Agents Details Collection Database Partition Activity Collection DB2 Partition Subcategories Wait Collection History Locks Wait Collection History Locks Summary Collection Instance Memory Collection Instance Memory Pool Collection Instance Memory Summary Collection Instance Memory Total Collection Database Memory Collection Database Memory Pool Collection Database Memory Pool DB Collection Database Memory Summary Collection Database Memory Total Collection Database Buffer Pools Collection Database Tables Collection Database Partition Backup Collection Database Tables Global Collection Database Applications Collection

Log Activity Collection

The following table provides a list of the Log Activity metrics that are collected, and a description for each.

 
Metric
Description

DB_NAME

The name of the monitored database.

DBPARTITIONNUM

The database partition from which the data was retrieved for this row.

TOTAL_LOG_AVAILABLE_KB

Use this element with TOTAL_LOG_USED_KB to determine whether you need to adjust the following configuration parameters to avoid running out of log space:

logfilsiz

logprimary

logsecond

TOTAL_LOG_USED_KB

The total amount of active log space currently used (in KB) in the database.

LOG_UTILIZATION

The percentage of used log within the entire available log.

SECONDARY_LOG_USED_TOP_KB

The maximum amount of secondary log space used (in KB).

The value is zero if the database does not have any secondary log files (which is the case if there are none defined).

TOTAL_LOG_USED_TOP_KB

The maximum amount of total log space used (in KB).

You can use this element to help evaluate the amount of primary log space that you have allocated. Comparing the value of this element with the amount of primary log space you have allocated can help you to evaluate your configuration parameter settings. Your primary log space allocation can be calculated using the following formula:

logprimary * logfilsiz * 4096

You can use this element with SECONDARY_LOG_USED_TOP_KB and SECONDARY_LOGS_ALLOCATED to evaluate your current dependency on secondary logs.

SECONDARY_LOGS_ALLOCATED

You may use this element with SEC_LOG_USED_TOP and TOT_LOG_USED_TOP to show your current dependency on secondary logs. If this value is consistently high, you may need larger log files, more primary log files, or more frequent COMMIT statements within your application.

As a result, you may need to adjust the following configuration parameters:

logfilsiz

logprimary

logsecond

logretain

LOG_READS

The number of log reads (pages) during the specified time range.

LOG_READS_RATE

The number of log reads (pages) per second during the specified time range.

LOG_READ_TIME_S

The total elapsed time spent by the logger reading log data from the disk.

LOG_WRITES

The number of log writes (pages) during the specified time range.

LOG_WRITES_RATE

The number of log writes (pages) per second during the specified time range.

LOG_WRITE_TIME_S

The total elapsed time spent by the logger writing log data to the disk.

NUMBER_LOG_WRITE_IO

The number of I/O requests issued by the logger for writing log data to the disk.

NUMBER_LOG_READ_IO

The number of I/O requests issued by the logger for reading log data from the disk.

NUMBER_LOG_PART_PAGE_IO

The number of I/O requests issued by the logger for writing partial log data to the disk.

NUMBER_LOG_BUFFER_FULL

The number of times agents that copy log records into the log buffer have to wait for log data to write to disk. This value is incremented each time an agent has to wait. For example, if two agents are waiting, the value is incremented by two.

NUMBER_LOG_DATA_FOUND_IN_
BUFFER

The number of times an agent reads log data from the buffer. Reading log data from the buffer is preferable to reading from the disk because the latter is slower.

Use this element with the NUM_LOG_READ_IO element to determine if the LOGBUFSZ database configuration parameter needs to be increased.

APPL_ID_OLDEST_XACT

The application ID (which corresponds to the AGENT_ID value from the application snapshot) of the application that has the oldest transaction.

This element can help in determining which application has the oldest active transaction. That application can be forced to free up log space. If it is taking up a great deal of log space, examine the application to determine if it can be modified to commit more frequently.

There are times when there is not a transaction holding up logging, or the oldest transaction does not have an application ID (for example, an in-doubt transaction or an inactive transaction). In these cases, an application ID is not returned in the data stream.

LOG_TO_REDO_FOR_RECOVERY_KB

The amount of log (in KB) that has to be redone for crash recovery.

When the snapshot is taken, this value is calculated based on the conditions at the time of the snapshot. Larger values indicate longer recovery times after a system crash. If the value seems excessive, check the LOG_HELD_BY_DIRTY_PAGES element to see if page cleaning needs to be tuned. Also, check if there are any long running transactions that need to be terminated.

LOG_HELD_BY_DIRTY_PAGES_KB

The amount of log corresponding to the difference between the oldest dirty page in the database and the top of the active log.

When the snapshot is taken, this value is calculated based on the conditions at the time of the snapshot.

Use this element to evaluate the effectiveness of page cleaning for older pages in the buffer pool.

The cleaning of old pages in the buffer pool is governed by the softmax database configuration parameter. If the page cleaning is effective, then LOG_HELD_BY_DIRTY_PAGES_KB should be less than or approximately equal to:

(softmax / 100) * logfilsiz * 4096. If this statement is not true, increase the number of page cleaners using the num_iocleaners configuration parameter.

If the condition is true and it is desired that less log be held by dirty pages, then decrease the softmax configuration parameter.

Log File Collection

The following table provides a list of the Log File metrics that are collected.

 
Metric
Description

INSTANCE

The name of the monitored instance.

FULL_DIAG_LOG_PATH

The full path and name of the diag log file (the filename is db2diag.log).

DIAG_LOG_SIZE

The diag log file size.

Log Message Collection

The following table provides a list of the Log Message metrics that are collected, and a description for each.

 
Metric
Description

APPL_ID

The application ID that is related to the message in the notification log file.

DBNAME

The name of the monitored database.

DBPARTITIONNUM

The name of the monitored database.

FUNCTION

The name of the function that generated the message.

This metric is retrieved for version 10 and higher.

IMPACT

Qualifies the impact of this message from a user's perspective. This clarifies the impact that the event described by the message has on the business process DB2 is part of. The impact is determined according to DB2 definition. The possible values are:

•
CRITICAL
•
IMMEDIATE
•
NONE
•
POTENTIAL
•
UNLIKELY

This metric is reserved for version 10 and higher.

MSG

The message text.

MSGNUM

The message number.

MSGSEVERITY

The message severity.

Values:

C (critical)

E (error)

W (warning)

I (informational)

NULL (if the message severity could not be determined)

MSGTYPE

The message type.

PID

The process ID that caused this message.

PROCESSNAME

The process name.

TID

The thread ID.

TIMESTAMP

The log message timestamp.

Messages Count Collection

The following table provides a list of the Messages Count metrics that are collected, and a description for each.

 
Metric
Description

DB_NAME

The name of the monitored database.

CRITICAL_COUNT

The number of critical messages for the selected period.

FATAL_COUNT

The number of fatal messages for the selected period.

WARNING_COUNT

The number of warning messages for the selected period.

INFORMATIONAL_COUNT

The number of informational messages for the selected period.

OTHER_COUNT

The number of unknown messages for the selected period.

相关文档

The document was helpful.

选择评级

I easily found the information I needed.

选择评级