Skip Headers
Oracle® Coherence Management Guide
Release 3.7.1

Part Number E22842-01
Go to Documentation Home
Home
Go to Book List
Book List
Go to Table of Contents
Contents
Go to Feedback page
Contact Us

Go to previous page
Previous
Go to next page
Next
View PDF

6 Analyzing Reporter Content

Coherence provides out-of-box reports that helps administrators and developers better analyze usage and configuration issues that may occur. For reports specific to Coherence*Web, see "Running Performance Reports" in Oracle Coherence User's Guide for Oracle Coherence*Web.

The following sections are included in this chapter:

6.1 Understanding the Cache Size Report

The cache size report can be executed either on demand or it can be added as part of the report batch and the Caches should have the <unit-calculator> subelement of <local-scheme> set to BINARY. The cache size file is a tab delimited file that is prefixed with the date and hour in YYYYMMDDHH format and post fixed with -cache-size.txt. For example 2009013101-cache-size.txt would be created on January 31, 2009 at 1:00 AM. Table 6-1 describes the content of the Cache Size report.

Table 6-1 Contents of the Cache Size Report

Column Data Type Description

Batch Counter

Long

A sequential counter to help integrate information between related files. This value does reset when the reporter restarts and is not consistent across members. However, it is helpful when trying to integrate files.

Cache Name

String

The name of the cache.

Cache Size

Double

The number of objects in the cache.

Memory Bytes

Double

The number of bytes consumed by the objects in the cache. This does not include indexes or over head.

MemoryMB

Double

The MB consumed by the objects in the cache. This does not include indexes or over head.

Avg Object Size

Double

The Average memory consumed by each object.


6.2 Understanding the Cache Usage Report

The cache usage report provides information about cache usage (gets, puts, evictions, and so on). The report is a tab-delimited file that is prefixed with the date and hour in YYYYMMDDHH format and post fixed with -cache-usage.txt. For example, 2010013113-cache-usage.txt would be created on January 31, 2010 1:00 pm. Table 6-2 describes the content of the cache utilization report.

Table 6-2 Contents of the Cache Usage Report

Column Data Type Description

Batch Counter

Long

A sequential counter to help integrate information between related files. This value resets when the Reporter restarts, and is not consistent across members. However, it is helpful when trying to integrate files.

Report Time

Date

The system time when the report is executed.

Service

String

The name of the cache service.

Cache Name

String

The name of the cache.

Tier

String

Value can be either front or back. Describes whether the cache resides in the front-tier (local cache) or back tier (remote cache).

Total Puts

Double

The total number of puts for the cache across the cluster since the last time the report was executed.

Total Puts Milliseconds

Double

The total number of milliseconds spent per put() invocation (PutsMillis) across the cluster since the last time the report was executed.

Total Gets

Double

The total number of gets for the cache across the cluster since the last time the report was executed.

Total Gets Milliseconds

Double

The total number of milliseconds spent per get() invocation (GetsMillis) across the cluster since the last time the report was executed.

Total Hits

Double

The total number of hits for the cache across the cluster since the last time the report was executed.

Total Hits Milliseconds

Double

The total number of milliseconds spent per get() invocation that is a hit (HitsMillis) across the cluster since the last time the report was executed.

Total Misses

Double

The total number of misses for the cache across the cluster since the last time the report was executed.

Total Misses Milliseconds

Double

The total number of milliseconds spent per get() invocation that is a miss (MissesMillis) across the cluster since the last time the report was executed.

Total Writes

Double

The total number of storage writes for the cache across the cluster since the last time the report was executed.

Total Writes Milliseconds

Double

The total number of milliseconds spent in storage write operations (WritesMillis) across the cluster since the last time the report was executed.

Total Reads

Double

The total number of reads from a cache store for the cache across the cluster since the last time the report executed.

Total Read Milliseconds

Double

The total number milliseconds on cache store reads for the cache across the cluster since the last time the report executed.

Total Failures

Long

The total number of storage failures for the cache across the cluster since the last time the report was executed.

Total Queue

Long

The sum of the queue link sizes across the cluster.

evictions

Long

The total number of evictions for the cache across the cluster since the last time the report was executed.

Cache Prunes

Long

The total number of prunes for the cache across the cluster since the last time the report was executed.

Cache Prunes Milliseconds

Long

The total number of milliseconds spent in the prune operation (PrunesMillis) across the cluster since the last time the report was executed.


6.3 Understanding the Flash Journal Report

The flash journal report displays statistics that are used to determine how well data is being stored to flash memory. The report is a tab-delimited file that is prefixed with the date and hour in YYYYMMDDHH format and post fixed with -flashjournal.txt. For example, 2010013113-flashjournal.txt would be created on January 31, 2010 1:00 pm. Table 6-3 describes the content of the cache utilization report.

Table 6-3 Contents of the Flash Journal Report

Column Type Description

Batch Counter

Long

A sequential counter to help integrate information between related files. This value does reset when the reporter restarts and is not consistent across members. However, it is helpful when trying to integrate files.

Report Time

Date

The system time when the report executed.

Node ID

Long

The member for the flash journal statistics.

FileCount

Integer

The number of journal files that are currently in use.

BinaryStoreCount

Integer

The number of active JournalBinaryStore objects that are using this journal.

TotalDataSize

Long

The amount of data, in bytes, that are currently stored for this journal.

TotalFileSize

Long

The total size of all journal files for this journal.

BacklogCount

Integer

The number of serialized values that have yet to be persisted.

BacklogSize

Integer

The maximum size, in bytes, of the backlog. The backlog is the amount of serialized values that has yet to be persisted. Client threads are blocked if this limit is exceeded and remain blocked until the backlog recedes below this limit

PoolSize

Integer

The total size, in bytes, of all available buffers in the pool.


6.4 Understanding the Management Report

The Management report contains refresh statistics that are used to determine if the management framework is providing a timely view of management data for all MBeans. The management file is a tab delimited file that is prefixed with the date and hour in YYYYMMDDHH format and post fixed with -Management.txt. For example 2009013113-Management.txt would be created on January 31, 2009 at 1:00 PM. Table 6-4 describes the content of the Management report.

Table 6-4 Contents of the Management Report

Column Type Description

Batch Counter

Long

A sequential counter to help integrate information between related files. This value does reset when the reporter restarts and is not consistent across members. However, it is helpful when trying to integrate files.

Report Time

Date

The system time when the report executed.

Refresh Policy

String

The refresh policy that is currently set, which is used to determine how data for remote models are being refreshed.

Refresh Time

Date

The timestamp when this model was last retrieved from a corresponding member. For local servers it is the local time.

Refresh Excess Count

Long

The number of times the MBean server predictively refreshed information and the information was not accessed.

Refresh Count

Long

The total number of snapshots retrieved since the statistics were last reset.

Refresh Prediction Count

Long

The number of times the MBean server used a predictive algorithm to refresh MBean information.

Refresh Timeout Count

Long

The number of times this management member has timed out while attempting to refresh remote MBean attributes.


6.5 Understanding the Memory Status Report

The Memory Status report must be run as part of a report batch. The values are helpful in understanding memory consumption on each member and across the grid. For data to be included, cluster members must be configured to publish platform MBean information. The memory status file is a tab delimited file that is prefixed with the date and hour in YYYYMMDDHH format and post fixed with -memory-status.txt. For example 2009013115-memory-status.txt would be created on January 31, 2009 at 3:00 PM. Table 6-5 describes the content of the Memory Status report.

Table 6-5 Contents of the Memory Status Report

Column Data Type Description

Batch Counter

Long

A sequential counter to help integrate information between related files. This value does reset when the reporter restarts and is not consistent across members. However, it is helpful when trying to integrate files.

Report Time

Date

The system time when the report executed.

JVM Uptime

Long

The amount of time since the JVM was started.

Node Id

Long

The member for the memory statistics.

Gc Name

String

The name of the Garbage Collector information.

CollectionCount

Long

The number of garbage collections that have happened since the computer started.

Delta Collection Count

Long

The number of garbage collections that have occurred since the last execution of the report.

CollectTime

Long

The number of milliseconds the JVM has spent on garbage collection since the start of the JVM.

Delta Collect Time

Long

The number of milliseconds the JVM has spent on garbage collection since the last execution of the report.

Last GC Start Time

Long

The start time of the last garbage collection.

Last GC Duration Millis

Long

The total amount of time of the last garbage collection.

Last GC Stop Time

Long

The stop time of the last garbage collection.

Heap Committed

Long

The number of heap bytes committed at the time of report.

Heap Init

Long

The number of heap bytes initialized at the time of the report.

Heap Max

Long

The Maximum number of bytes used by the JVM since the start of the JVM.

Heap Used

Long

The bytes used by the JVM at the time of the report.


6.6 Understanding the Network Health Detail Report

The Network Health report supporting member-level details for determining the health of the network communications. The network health detail file is a tab delimited file that is prefixed with the date and hour in YYYYMMDDHH format and post fixed with -network-health-detail.txt. For example 2009013114-network-health.txt would be created on January 31, 2009 at 2:00PM. Table 6-6 describes the content of the Network Health Detail report.

Table 6-6 Contents of the Network Health Detail Report

Column Data Type Description

Batch Counter

Long

A sequential counter to help integrate information between related files. This value does reset when the reporter restarts and is not consistent across members. However, it is helpful when trying to integrate files.

Report Time

Date

The system time when the report executed.

RefreshTime

Date

The system time when management information was last retrieved from a corresponding node. Local servers display the local time.

Node Id

Long

The member for the network statistics.

Tx Success

Double

The publisher success rate for the member. If this value is within 2%-3% of the "Min Node Tx Success" and more than 10% less than the "Grid Tx Success" for the batch in the Network Health File, the corresponding member may be having difficulty communicating with the cluster. Constrained CPU, constrained network bandwidth or high network latency could cause this to occur.

RX Success

Double

The receiver success rate for the member. If this value is within 2%-3% of the "Min Node Rx Success" and more than 10% less than the "Grid Tx Success" for the batch in the Network Health File, the corresponding member may be having difficulty communicating with the cluster. Constrained CPU, constrained network bandwidth or high network latency could cause this to occur.

Packets Sent

Double

The total number of network packets sent by the member.

Current Packets Sent

Long

The number of packets sent by the member since the prior execution of the report.

Packets Resent

Long

The total number of network packets resent by the member. Packets are resent when the receiver of the packet receives and invalid packet or when an acknowledge packet is not sent within the appropriate amount of time.

Current Packet Resent

Long

The number of network packets resent by the member since the prior execution of the report.

PacketsRepeated

Long

The total number of packets received more than once.

Current Packets Repeated

Long

The number of packets received more than once since the last execution of the report.

Packets Received

Long

The total number of packets received by the member.

Current Packets Received

Long

The total number of packets received by the member since the last execution of the report.


6.7 Understanding the Network Health Report

The Network Health report contains the primary aggregates for determining the health of the network communications. The network health file is a tab delimited file that is prefixed with the date and hour in YYYYMMDDHH format and post fixed with -network-health.txt. For example 2009013113-network-health.txt would be created on January 31, 2009 at 1:00 PM. Table 6-7 describes the content of the Network Health report.

Table 6-7 Contents of the Network Health Report

Column Type Description

Batch Counter

Long

A sequential counter to help integrate information between related files. This value does reset when the reporter restarts and is not consistent across members. However, it is helpful when trying to integrate files.

Report Time

Date

The system time when the report executed.

Min Node Rx Success

Double

The minimum receiver success rate for a member in the cluster. If this value is considerably less (10%) than the Grid Rx Success rate. Further analysis using the Network Health Detail should be done.

Grid Rx Success

Double

The receiver success rate for the grid as a whole. If this value is below 90%. Further analysis of the network health detail should be done.

Min Node Tx Success

Double

The minimum publisher success rate for a member in the cluster. If this value is considerably less (10%) than the Grid Rx Success rate. Further analysis using the Network Health Detail should be done.

Grid TX Success

Double

The publisher success rate for the grid as a whole. If this value is below 90%. Further analysis of the network health detail should be done.


6.8 Understanding the Node List Report

Due to the transient nature of the node identifier (nodeId), the reporter logs out a list of members and user-defined member identity information. See "member-identity" in the Oracle Coherence Developer's Guide. The node list file is a tab delimited file that is prefixed with the date and hour in YYYYMMDDHH format and post fixed with -nodes.txt. For example 2009013101-nodes.txt would be created on January 31, 2009 at 1:00 AM. Table 6-8 describes the content of the Node List report.

Table 6-8 Contents of the Node List Report

Column Data Type Description

Batch Counter

Long

A sequential counter to help integrate information between related files. This value does reset when the reporter restarts and is not consistent across members. However, it is helpful when trying to integrate files.

Report Time

Date

The system time when the report executed.

Node Id

String

The numeric member identifier.

Unicast Address

String

The Unicast address for the member.

Member Name

String

The member name.

Process Name

String

The process name for the member.

Role Name

String

The role name for the member.

Machine Name

String

The computer name for the member.

Rack Name

String

The rack name for the member.

Site Name

String

The site name for the member.

Refresh Time

Date/Time

The time which the information was refreshed from a remote member. If the time is different than the refresh time on other rows in the batch, the member did not respond in a timely matter. This is often caused by a member preforming a garbage collection. Any information regarding a member with an "old" refresh date is questionable.


6.9 Understanding the Proxy Report

The proxy file provides information about proxy servers and the information being transferred to clients. The Proxy file is a tab delimited file that is prefixed with the date and hour in YYYYMMDDHH format and post fixed with -report-proxy.txt. For example 2009013101-report-proxy.txt would be created on January 31, 2009 at 1:00 AM. Table 6-9 describes the content of the Proxy report.

Table 6-9 Contents of the Proxy Report

Column Type Description

Batch Counter

Long

A sequential counter to help integrate information between related files. This value does reset when the reporter restarts and is not consistent across members. However, it is helpful when trying to integrate files.

Report Time

Date

The system time when the report executed.

Node Id

String

The numeric member identifier.

Service Name

String

The name of the proxy service.

HostIp

String

The IP Address and Port of the proxy service.

Connection Count

Long

The current number of connections to the proxy service.

Outgoing Byte Backlog

Long

The number of bytes queued to be sent by the proxy service.

Outgoing Message Backlog

Long

The number of messages queued by the proxy service.

Bytes Sent

Long

The number of bytes sent by the proxy service since the last execution of the report.

Bytes Received

Long

The number of bytes received by the proxy service since the last execution of the report.

Messages Sent

Long

The number of messages sent by the proxy service since the last execution of the report.

Messages Received

Long

The number of messages received by the proxy service since the last execution of the report.


6.10 Understanding the Ram Journal Report

The ram journal report displays statistics that are used to determine how well data is being stored to RAM memory. The report is a tab-delimited file that is prefixed with the date and hour in YYYYMMDDHH format and post fixed with -ramjournal.txt. For example, 2010013113-ramjournal.txt would be created on January 31, 2010 1:00 pm. Table 6-10 describes the content of the cache utilization report.

Table 6-10 Contents of the Ram Journal Report

Column Type Description

Batch Counter

Long

A sequential counter to help integrate information between related files. This value does reset when the reporter restarts and is not consistent across members. However, it is helpful when trying to integrate files.

Report Time

Date

The system time when the report executed.

Node ID

Long

The member for the flash journal statistics.

FileCount

Integer

The number of journal files that are currently in use.

BinaryStoreCount

Integer

The number of active JournalBinaryStore objects that are using this journal.

TotalDataSize

Long

The amount of data, in bytes, that are currently stored for this journal.

TotalFileSize

Long

The total size of all journal files for this journal.


6.11 Understanding the Service Report

The service report provides information to the requests processed, request failures, and request backlog, tasks processed, task failures and task backlog. Request Count and Task Count are useful to determine performance and throughput of the service. RequestPendingCount and Task Backlog are useful in determining capacity issues or blocked processes. Task Hung Count, Task Timeout Count, Thread Abandoned Count, Request Timeout Count are the number of unsuccessful executions that have occurred in the system. Table 6-11 describes the contents of the Service report.

Table 6-11 Contents of the Service Report

Column Data Type Description

Batch Counter

Long

A sequential counter to help integrate information between related files. This value does reset when the reporter restarts and is not consistent across members. However, it is helpful when trying to integrate files.

Report Time

Date

The system time when the report executed.

Service

String

The service name.

Node Id

String

The numeric member identifier.

Refresh Time

Date

The system time when the service information was updated from a remote member.

Request Count

Long

The number of requests since the last report execution.

RequestPendingCount

Long

The number of pending requests at the time of the report.

RequestPendingDuration

Long

The duration for the pending requests at the time of the report.

Request Timeout Count

Long

The number of request timeouts since the last report execution.

Task Count

Long

The number of tasks executed since the last report execution.

Task Backlog

Long

The task backlog at the time of the report execution.

Task Timeout Count

Long

The number of task timeouts since the last report execution.

Task Hung Count

Long

The number of tasks that hung since the last report execution.

Thread Abandoned Count

Long

The number of threads abandoned since the last report execution.