7.3.3 Collecting failure information (for prerequisite products other than a service platform)
This section outlines the procedures for collecting the failure information that is output by major prerequisite products other than a service platform. For details about the failure information that is output and how to make estimates about log files and trace files, see the applicable manual for the relevant product.
- Organization of this subsection
(1) Collecting the failure information that is output by the J2EE server
The default directory for the logs output by the J2EE server is <Installation directory of Service Platform>\CC\server\public\ejb\<J2EE server name>\logs. This directory is mentioned as "Log output directory of J2EE server" hereafter. You can change the directory of the log output by the J2EE server with the ejb.server.log.directory key of usrconf.cfg (option definition file for J2EE server).
(a) Types of failure information and how to collect it
When a failure occurs and it is difficult to specify the cause of failure from the log or the trace output by the Service Platform, the following failure information output by the J2EE server is required:
-
Log of J2EE servers, redirectors, and server management commands
-
Log of Administration Agent, Management Agent, and Management Server
-
DB Connector log
-
Log of EJB client application
-
User log
-
Maintenance log
Use the following commands to acquire the failure information output by the J2EE server in a batch (can be acquired manually). When failure information is acquired in a batch, there may be information that partially overlaps with the log and the trace output by the Service Platform. For details on how to acquire the failure information output by the J2EEserver, see "Application Server Maintenance and Migration Guide".
-
snapshotlog command
For details on the snapshot log command, see "snapshotlog (collect snapshot logs)" in "Application Server Command Reference Guide".
-
mngsvrutil command (specify collect snapshot in the sub command)
For details on the mngsvrutil command, see "mngsvrutil (Management Server management command)" in "Application Server Command Reference Guide".
By executing the above commands, you acquire the information in the following directories:
-
<Installation directory of Service Platform>\CC\server\public\ejb\*\logs\csc
-
<Installation directory of Service Platform>\CC\server\public\ejb\*\logs\csc\maintenance
-
<Installation directory of Service Platform>\CSC\config\manager
-
<Installation directory of Service Platform>\CSC\config\msg
-
<Installation directory of Service Platform>\CSC\log\manager
-
<Installation directory of Service Platform>\CSC\repository
-
<Installation directory of Service Platform>\CSC\spool
-
<Installation directory of Service Platform>\CSC\system\msg
Specify the information to be acquired in the definition file for collecting a snapshot log. If required, change the definition file used for collecting a snapshot log, and then change the information to be acquired. For details on the definition file for collecting a snapshot log, see "10.2.1 Definition file for snapshot log collection" in "Application Server Definition Reference Guide".
(b) Preparation before start of operation
Define the type (Path) of the failure information to be acquired beforehand when you perform batch acquisition by using commands. Specify the primary delivery material that is sent immediately by mails and the secondary delivery material that is sent separately. For details, see "2.4 Types of Required Data" in "Application Server Maintenance and Migration Guide".
(c) Definition files required for failure investigation
The definitions set up in the J2EE server are required for investigating failure. The following table describes definition files other than the Service Platform:
|
Definition file |
Description |
|---|---|
|
Setup file of Component Container#1 |
All the files and directories under <Installation directory of Service Platform>\CC\server\usrconf\ejb\<J2EE server name>\directory |
|
Setup file of Manager#1 |
All the files and directories under <Installation directory of Service Platform>\manager\config\directory |
|
Property information of DB Connector set in dbcon-xadisplayname #2 |
Definition file set in "3.1.2(8) Setting up DB Connector(When Reliable Messaging is used)". |
|
Property information of DB Connector set in dbcon-nodisplayname#2 |
- Note#1
-
For details on definitions, see "Application Server Definition Reference Guide ".
- Note#2
-
For details on definitions, see "4.1 HITACHI Connector Property file" in "Application Server Application and Resource Definition Reference Guide".
(2) How to acquire the failure information output by Reliable Messaging
You can identify the cause of failure by acquiring the failure information output by Reliable Messaging when using the following reception or service adapter:
-
Standard reception of asynchronous (MDB (WS-R))
-
Standard reception of asynchronous (MDB (database queue))
-
MDB (WS-R) adapter
-
MDB (Database queue) adapter
For details on how to acquire failure information output by Reliable Messaging, see "9.1 Overview of the information output at the time of failure" in "Reliable Messaging".
(a) Messages output by Reliable Messaging
The messages output by Reliable Messaging are output to the following log. There are 2 types of logs; start and stop messages log and message log for the J2EE server. The following table lists and describes the messages output by Reliable Messaging:
|
Log type |
Output contents |
Directory and file name |
|---|---|---|
|
Start and stop messages log |
Information on start and stop, and value of the property specified by the user. |
|
|
Message log for the J2EE server |
Information of Reliable Messaging for starting, stopping, and during operations. |
|
(b) Relationship between the HCSC server and the trace of Reliable Messaging
The relationship between the HCSC server and the trace of Reliable Messaging differs when you use the local queue and the transfer queue, and when you use the shared queue.
-
When you use the local queue or the transfer queue
There are two types of Reliable Messaging trace namely; method trace and line trace. The method trace outputs error information (cause of exception occurrence) and the entry information, exit information, and debug information of method. The line trace outputs the information of messages transmitted between queues. The following figure shows the trace collection point when using local queue or transfer queue:
Figure 7‒2: Trace collection point when using local queue or transfer queue The following table lists and describes the relationship between trace collection point and file names. "Number in figure" in this table corresponds to the numbers in "Figure 7-2 Trace collection point when using local queue or transfer queue".
Table 7‒7: Relationship between the trace collection point and file names Number in figure
Collection point
Trace
Type
Output machine
Output destination
File name
1
When sending a request from the service requester to transfer queue
Method trace
Service requester machine
<Log output directory of the J2EE server>\RM\maintenance
mtd_<Display name of Reliable Messaging of the service requester>_<numbers>.log
2
When transferring from the transfer queue of the service requester to the ocal queue of HCSC server
Line trace
lin_<Display name of Reliable Messaging of the service requester>_<numbers>.log
3
When accepting transfer from the transfer queue of the service requester to the local queue of the HCSC server, and when transferring from the transfer queue of the HCSC server to the local queue of the service component.
Line trace
Machine on which the HCSC server is running
<Log output directory of the J2EE server >\RM\maintenance
lin_<Display name of Reliable Messaging of the HCSC server>_<numbers>.log
4
When HCSC server fetches messages from the local queue, and when sending requests from the service adapter to the transfer queue
Method trace
mtd_<Display name of Reliable Messaging of the HCSC server>_<numbers>.log
5
When accepting transfer from the transfer queue at the HCSC server to the local queue of the service component
Line trace
Machine on which the service component is running
<Log output directory of the J2EE server >\RM\maintenance
lin_<Display name of Reliable Messaging of the service component >_<numbers>.log
6
When the service component fetches messages from the local queue
Method trace
mtd_<Display name of Reliable Messaging of the service component >_<numbers>.log
-
When you use the shared queue
There are two types of Reliable Messaging trace namely, method trace and shared queue event trace. The method trace outputs the error information (cause for occurrence of exception) and entry information, exit information, and debug information of the method. The share queue event trace outputs the information of an event indicating that the messages are stored in share queue. The following figure shows the trace collection point when using the shared queue:
Figure 7‒3: Trace collection point when using the shared queue The following table lists and describes the relationship between the trace collection point and the file name. "Number in figure" in this table corresponds to the numbers in "Figure 7-3 Trace collection point when using the shared queue".
Table 7‒8: Relationship between the trace collection point and the file name Number in figure
Collection point
Trace
Type
Output machine
Output destination
File name
1
When messages are saved from the service requester to the shared queue for reception in HCSC server, and when messages are saved from the service adapter of HCSC server to the shared queue for sending
Shared queue event trace
Machine on which the HCSC server is running
<Log output directory of the J2EE server >\RM\maintenance
shq_<Display name of Reliable Messaging of the HCSC server>_<numbers>.log
2
When HCSC server fetches messages from the local queue, and when sending requests from service adapters to transfer queue
Method trace
mtd_<Display name of Reliable Messaging of the HCSC server >_<numbers>.log
(c) Preparation before start of operation
To change the type of information that is output by changing the output level of method trace or line trace, you must change Reliable Messaging settings before starting the HCSC server. Perform these changes by using the RMMethodTraceLevel and RMLineTraceLevel configuration properties of Reliable Messaging.
(d) Definition files required for failure investigation
The information set up for Reliable Messaging configuration properties is required for investigating failures. The following table describes the definition files of Reliable Messaging required for investigating failures. For details on definition files, see " 6. Configuration Property" in "Reliable Messaging".
|
Definition file |
Description |
|---|---|
|
Setup information of Reliable Messaging properties |
Definition files setup in "3.1.2(9) Setting up Reliable Messaging". |
(3) How to acquire the failure information output by SOAP Communication Infrastructure or JAX-WS engine
When using synchronous standard reception(Web Service) or SOAP adapter, you can identify the cause of failure by acquiring the failure information output by SOAP Communication Infrastructure or the JAX-WS engine. You can identify the cause of failure when there is an error in the sent and received user messages, by checking the message format of SOAP Messages when invoking the HCSC server from service requester, and the SOAP Messages when invoking service component from the service adapter of the HCSC server.
For details on how to acquire the failure information output by SOAP Communication Infrastructure ,see "14.Troubleshooting" in "Application Server SOAP Application Development Guide".
For details on how to acquire the failure information output by JAX-WS engine, see "39.Troubleshooting" in "Application Server Web Service Development Guide".
(a) Relationship between the HCSC server and SOAP trace or Application log
This sub-section describes the relationship between the collection point and SOAP trace or file names of application logs in the HCSC server configuration. This relationship differs when invoking SOAP adapter from the reception of Web Services and when invoking SOAP adapter from a reception other than Web Services.
-
When invoking SOAP adapter from synchronous standard reception (Web Services) or SOAP reception
The SOAP trace or application log is output with the following names when you invoke the SOAP adapter from synchronous standard reception (Web Services) or from the SOAP reception. The same names are output when it passes through the business process on the way. The following figure shows the collection point of SOAP trace and application log:
Figure 7‒4: Collection point of SOAP trace or application logs (when you invoke the SOAP adapter from the reception of Web Services) The following table lists and describes the relationship between the trace collection point and file names. The numbers in the Number in figure column of the table correspond to the numbers in the above figure.
Table 7‒10: Relationship between trace collection points and file names (when you invoke the SOAP adapter from the reception of Web Services) Number in figure
Collection point
File name
Trace#1
Application log#1
1#2
SOAP communication trace when requests (sending) are sent from the service requester, and when there is a response (receiving) from the service requester
- When using SOAP Communication Infrastructure
-
"<prefix>-j2ee-<J2EE server name>-<numbers>.log"
<prefix>#3 is set in the client definition file of service requester side.
- When using JAX-WS engine
-
"cjwmessage<numbers>.log"
- When using SOAP Communication Infrastructure
-
"<prefix>-j2ee-<J2EE server name>-aplog-<numbers>.log"
<prefix>#3 is set in the client definition file of service requester side.
- When using JAX-WS engine
-
"cjwtransport<numbers>.log"
2
SOAP Communication trace when requests are received (receiving) from the service requester with the request reception of the HCSC server, and when there is a response (sending) from the HCSC server
- For synchronous standard reception (Web Services)
-
"<cluster name>-j2ee-<J2EE server name>-<numbers>.log"
- For SOAP reception
-
"<Reception ID>-j2ee-<J2EE server name>-<numbers>.log"
- For synchronous standard reception (Web Services)
-
"<cluster name>-j2ee-<J2EE server name>-aplog-<numbers>.log"
- For SOAP reception
-
"<Reception ID>-j2ee-<J2EE server name>-aplog-<numbers>.log"
3
SOAP Communication trace when requests are sent (sending) from the service adapter of the HCSC server, and when there is a response (receiving) from the service adapter
Output to a trace file same as number 2 in the figure.
Output to the application log file same as number 2 in the figure.
4#2
Communication trace when requests from the service adapter of the HCSC server are received (receiving) in the service component , and when there is a response (sending) from the service component
- When using SOAP Communication Infrastructure
-
"<prefix>-j2ee-<J2EE server name>-<numbers>.log"
<prefix>#3 is set in the server definition file of the service component.
- When using JAX-WS engine
-
"cjwmessage<numbers>.log"
- When using SOAP Communication Infrastructure
-
"<prefix>-j2ee-<J2EE server name>-aplog-<numbers>.log"
<prefix>#3 is set in the server definition file of the service component.
- When using JAX-WS engine
-
"cjwmessage<numbers>.log"
- Note#1
-
The output destinations of trace and application logs differ as follows according to the Communication Infrastructure in use:
When using SOAP Communication Infrastructure
The <Log output directory of J2EE server>\WS on each running machine.
When using JAX-WS engine
The < Log output directory of J2EE server>\CJW on each running machine.
- Note#2
-
Traces can be collected when the SOAP Communication Infrastructure or JAX-WS engine provided by Service Platform is used on the machine on which the service requester or service component is running.
- Note#3
-
The name setup in prefix (c4web.logger.<identifier>.log_file_prefix) of trace files or application logs in the service definition file (c4websv.cfg) or the client definition file (c4webcl.properties).
-
When invoking SOAP adapters from a reception other than Web Services
The SOAP trace or application log is output with the following names when you invoke SOAP adapters from a request reception other than the synchronous standard reception or the SOAP reception. The following figure shows the collection point of SOAP trace and application log:
Figure 7‒5: Collection points of SOAP trace or application log (when you invoke a service adapter from a reception other than Web Services) The following table lists and describes the relationship between the trace collection point and file names. The numbers in the Number in figure column of the table correspond to the numbers in the above figure.
Table 7‒11: Relationship of trace collection point and file names (when you invoke a SOAP adapter from a reception other than Web Services) Number in figure
Collection point
File name
Trace#1
Application log#1
1
SOAP Communication trace when requests are sent (sending) from the service adapters of the HCSC server, and when there is a response (receiving) from service adapters
"<service ID>-j2ee-<J2EE server name>-<numbers>.log"#3
"<service ID>-j2ee-<J2EE server name>-aplog-<numbers>.log"#3
2#2
SOAP Communication trace when requests from the service adapter of the HCSC server are received (receiving) by service components, and when there is a response (sending) from service components
- When using SOAP Communication Infrastructure
-
"<prefix>-j2ee-<J2EE server name>-<numbers>.log"
<prefix>#4 is set in the server definition file of the service component.
- When using JAX-WS engine
-
"cjwmessage<numbers>.log"
- When using SOAP Communication Infrastructure
-
"<prefix>-j2ee-<J2EE server name>-aplog-<numbers>.log"
<prefix>#4 is set in the server definition file of the service component.
- When using JAX-WS engine
-
"cjwtransport<numbers>.log"
- Note#1
-
The output destination of trace and application logs differs as follows according to the Communication Infrastructure in use:
When using SOAP Communication Infrastructure
The <Log output directory of J2EE server>\WS on each running machine.
When using JAX-WS engine
The <Log output directory of J2EE server>\CJW on each running machine.
- Note#2
-
Can be acquired when a machine on which the service component is running uses SOAP Communication Infrastructure or the JAX-WS engine provided by the Service Platform.
- Note#3
-
Even if you specify prefix of trace files and application logs in the client definition file setup in the Service adapter definition screen, the specified value is not enabled. The service ID is assigned as a unique prefix.
- Note#4
-
The name where the prefix (c4web.logger.<identifier>.log_file_prefix) is setup for trace files or application logs in the server definition file (c4websv.cfg).
(b) Preparation before start of operation
If you acquire the log of SOAP messages, you can identify the cause of failure when there is an error in the sent and received user messages by checking the message format of SOAP Messages when invoking the HCSC server from the service requester and the SOAP messages when invoking the service component from the service adapter of the HCSC server.
The log output is set in the following common definition files on which HCSC server is running:
-
For SOAP Communication Infrastructure: c4webcom.cfg
-
For JAX-WS engine:cjwconf.properties
The following table describes the settings to the common definition file of the machine on which HCSC server is running:
|
Item |
Key name |
Setting (Default) |
Description |
|---|---|---|---|
|
Importance of application log output |
c4web.logger.aplog_level |
WARN |
Specify INFO while capturing (check the message format of SOAP Messages) the log of SOAP Messages. Must be set in advance before starting the HCSC server. |
For details on creating common definition files used by SOAP Communication Infrastructure, see"10.4 Setting up the common definition file" in "Application Server SOAP Application Development Guide".
For details on creating common definition files used by the JAX-WS engine, see"10.1.2 Settings of the common definition file" in "Application Server Web Service Development Guide".
(c) Definition files required for failure investigation
The setup information of SOAP Communication Infrastructure and the JAX-WS engine is required for investigating failures. The following table describes definition files of SOAP Communication Infrastructure:
|
Definition file |
Description |
File save location |
File name |
|---|---|---|---|
|
Server definition file |
Controls the operations of SOAP Communication in the request reception of a machine on which the HCSC server is running |
Following folders of machine on which the HCSC server is running:
|
c4websv.cfg |
|
Common definition file |
Controls the common operations of SOAP Communication the in request reception and the service adapter of a machine on which HCSC server is running |
Following folders of machine on which the HCSC server is running:
|
c4webcom.cfg |
For details on definition files when using SOAP Communication Infrastructure, see "10.2 Setting up the server definition file" in "Application Server SOAP Application Development Guide". For details on common definition files when using SOAP Communication Infrastructure, see "10.4 Setting the common definition file" in "Application Server SOAP Application Development Guide ".
The following table describes the definition file for the JAX-WS engine:
|
Definition file |
Explanation |
File storage location |
File name |
|---|---|---|---|
|
Common definition file |
Controls the SOAP communication operations common to the request reception and the service adapter of a machine on which the HCSC server is running. |
Following folders of machine on which the HCSC server is running:
|
cjwconf.properties |
For details on definition files when using JAX-WS engine, see "10.1 Action definition file" in "Application Server Web Service Development Guide".
(4) How to acquire the failure information output by the HTTP Server
The failure information output by the HTTP Server is required when operations are performed using the HTTP Server. For details on how to acquire the failure information output by HTTP Server, see "HTTP Server".
(5) How to acquire failure information output by database
When the cause of failure exists in the database, you can identify the cause by checking the failure information of the database. There is server side and client side information in the failure information of database. The following figure shows the relationship between the database (client side) and of a machine on which HCSC server is running and the database (server side):
|
|
|
Database used |
Server/Client |
Failure information |
|---|---|---|
|
When HiRDB is used |
Server side failure information |
|
|
Server side failure information (definition information) |
|
|
|
Client side failure information |
|
|
|
When Oracle is used |
Server side failure information |
Oracle log |
|
Client side failure information |
Oracle JDBC Thin Driver log |
- Note#
-
For HiRDB Type4 JDBC Driver, the exception trace log is disabled even if it is described in the environment variable of OS or HiRDB.ini OS. Specify by connecting with semicolons in the property definition "EnvironmentVariables" of the used DB Connector as follows:
(Example) PDSQLTRACE=0;PDCLTPATH=c:\tmp
If you use HiRDB, see " HiRDB System Operation Guide". If you use Oracle, see the documentation published by Oracle Corporation.
(6) How to acquire JavaVM information
(a) JavaVM output message log
When the JavaVM crashes, the debug information is output to standard output and error report file by JavaVM. For details, see "4.11 JavaVM Output Message Logs (Standard Output or Error Report File)" in "Application Server Maintenance and Migration Guide". The output destination and output file name of the error report file is as follows:
-
For Windows
<Installation directory of Service Platform>\CC\server\usrconf\ejb\<J2EEserver name>\hs_err_pid<Process ID of server process>.log
-
For UNIX
<Installation directory of Service Platform>/CC/server/usrconf/ejb/<J2EE server name>/hs_err_pid<Process ID of server process>.log
(b) Thread dump of JavaVM
JavaVM statistics and stack status of threads is output. For details, see "4.7 JavaVM thread dump" in "Application Server Maintenance and Migration Guide". The output destination is as follows:
-
In Windows
<Installation directory of Service Platform>\CC\server\usrconf\ejb\<J2EE server name>\javacore*.txt
-
In UNIX
<Installation directory of Service Platform>/CC/server/usrconf/ejb/<J2EEserver name>/javacore*.txt
(c) Java VM GC log
The activities of Java VM GC are output to this log. For details, see "4.8 JavaVM GC Log" in "Application Server Maintenance and Migration Guide".
(d) Memory dump
When there is a failure in the system, the system administrator or Maintenance Service use the memory dump to analyze the error. For details, see "4.9 Memory Dump" in "Application Server Maintenance and Migration Guide".