HiCommand Dynamic Link Manager User's Guide (for Windows(R))

3.7.2 Setting the HDLM functionality by using the Options window and the set operation

The recommended and default values for the HDLM functionalities are shown below in Table 3-11 The recommended and default values for each function.

Table 3-11 The recommended and default values for each function

Functionality Default value Recommended value

Load-balancing ON
Algorithm is round robin ON
The recommended algorithm depends on the operating environment.

Path health checking ON
30-minute check interval ON
The recommended checking interval depends on the operating environment.

Automatic failback OFF OFF

Intermittent Error Monitor OFF OFF

Remove LU OFF The recommended value depends on the operating environment.

Logging level 3: Collect all information for errors with the level "Information" or higher 3: Collect all information for errors with the level "Information" or higher

Trace level 0: Do not output a trace 0: Do not output a trace

File size (for Error log) 9900 (KB) 9900 (KB)

Number of files (for Error log) 2 2

File size (for Trace) 1000 (KB) 1000 (KB)

Number of files (for Trace) 4 4

Legend:

For the dlnkmgr command's set operation, enter lower-case characters such as on and off, instead of ON and OFF, as shown in the table.

Organization of this subsection

(1) Setting load balancing

(2) Setting path health checking

(3) Setting automatic failback

(4) Settings for intermittent error monitoring

(5) Setting the reservation level

(6) Setting the LU deletion functionality

(7) Setting the error log collection level

(8) Setting the trace level

(9) Setting the error log file size

(10) Setting the Number of Error Log Files

(11) Setting the trace file size

(12) Setting the number of trace files

Functionality	Default value	Recommended value
Load-balancing	ON Algorithm is round robin	ON The recommended algorithm depends on the operating environment.
Path health checking	ON 30-minute check interval	ON The recommended checking interval depends on the operating environment.
Automatic failback	OFF	OFF
Intermittent Error Monitor	OFF	OFF
Remove LU	OFF	The recommended value depends on the operating environment.
Logging level	3: Collect all information for errors with the level "Information" or higher	3: Collect all information for errors with the level "Information" or higher
Trace level	0: Do not output a trace	0: Do not output a trace
File size (for Error log)	9900 (KB)	9900 (KB)
Number of files (for Error log)	2	2
File size (for Trace)	1000 (KB)	1000 (KB)
Number of files (for Trace)	4	4

(1) Setting load balancing

You can select whether to enable load balancing. In a cluster environment, the load balancing function is only available for Thunder 9200/9500V Series, Lightning 9900/9900V Series, TagmaStore, and Universal Storage Platform V. In a non-cluster environment, the load balancing function is available for Thunder 9200/9500V Series, Lightning 9900/9900V Series, TagmaStore, Universal Storage Platform V, and the EMC DMX series. The above-mentioned conditions are applied regardless of the settings for the EMC DMX series, EMC CX series ,and HP EVA series. For example, in a cluster environment, the load balancing function is disabled even if it is set to be enabled for the EMC DMX series. For details on the cluster software supported by HDLM, see 2.12 Cluster support.
Setting by using the Options window
Set it to ON to enable load balancing. Otherwise, set it to OFF. After selecting ON, select a desired algorithm to be used for load balancing from the pull-down list.
Setting by the set operation
The following is an example of a command to set load balancing.
 
>dlnkmgr set -lb on -lbtype rr
 
Set it to on to enable load balancing. Otherwise, set it to off. When you set this value to on, specify rr for round robin or exrr for extended round robin after the -lbtype option. The type of algorithm specified by the -lbtype parameter remains stored in the system, even when you disable the load balancing function by specifying -lb off. Therefore, when you re-enable load balancing without specifying an algorithm, load balancing will be executed according to the settings stored in the system.

(2) Setting path health checking

You can select whether to enable path health checking.
Setting by using the Options window
Set it to ON to enable path health checking. Otherwise, set it to OFF. When selecting ON, you can specify Checking interval for path health checking.
Setting by using the set operation
The following is an example command to set path health checking.
 
>dlnkmgr set -pchk on -intvl 10
 
To enable path health checking, set it to on. Otherwise, set it to off. When you set this value to on, you can use the -intvl parameter to specify the checking interval. The value that was specified previously will be applied if the checking interval is not specified. For example, specify the path health check as off after specifying the checking interval as 15 minutes and executing. Then, when executing after specifying the path health check as on without specifying the checking interval, the 15 minutes that were specified previously will be applied again.

(3) Setting automatic failback

You can select whether to enable automatic failback.

When intermittent error monitoring is enabled and the number of error occurrence is 2 or more, the following condition must be satisfied.
 
error-monitoring-interval >= checking-interval-for-automatic-failback x number-of-times-error-is-to-occur-during-intermittent-error-monitoring
 
If this condition is not satisfied, an error occurs and the warning message (KAPL02064-W for the Options window and KAPL01080-W for the set operation) is output.

In such a case, change any of the following settings: the checking interval for automatic failback, intermittent error monitoring interval, or the number of times that the error is to occur.

When you set the number of times that the error is to occur to 1, the condition above does not need to be satisfied.
Setting by using the Options window
To enable automatic failback, set it to ON. Otherwise, set it to OFF. When you specify ON, you can select the checking interval for automatic failback.
Setting by using the set operation
The following shows an example of executing the command to set automatic failback:
 
>dlnkmgr set -afb on -intvl 10
 
To enable automatic failback, set it to on . Otherwise, set it to off. The checking interval can be specified with the -intvl parameter when on is specified. The value that was specified previously will be applied if the checking interval is not specified. For example, specify auto failback as off after specifying the checking interval to five minutes and executing. Then, when executing after specifying auto failback as on without specifying the checking interval, the five minutes that were specified previously will be applied again.

(4) Settings for intermittent error monitoring

Intermittent error monitoring is specifiable only when automatic failback is enabled. To prevent an intermittent error from reducing I/O performance, we recommend that you monitor intermittent errors when automatic failback is enabled.

When intermittent error monitoring is enabled, you can specify intermittent error conditions (the conditions used by the system to determine whether an intermittent error is occurring). The default value for the intermittent error monitoring interval is 30. The default value for the number of error occurrences is 3.

The system assumes that an intermittent error is occurring if the specified number of times that the error is to occur is reached during the monitoring interval (from the time that the monitoring interval starts, until the specified interval ends). A path that is assumed to have an intermittent error is excluded from automatic failback. Intermittent error monitoring starts when the path is recovered from the error by using automatic failback. Monitoring is performed on individual paths.

When a value of 2 or more is specified in number of times, make sure that the condition shown in (3) Setting automatic failback is satisfied.

To determine whether a path is ineligible for automatic failback, you can use the results of the dlnkmgr command's view operation or the Path List view of the HDLM GUI.
Setting by using the Options window
To enable intermittent error monitoring, specify ON. To disable intermittent error monitoring, specify OFF. When you specify ON, you can specify intermittent error conditions (the conditions used by the system to determine whether an intermittent error is occurring) in Monitoring interval and Number of times.
Setting by using the set operation
The following shows an example of executing the command to set the error log collection level.
 
>dlnkmgr set -iem on -intvl 20 -iemnum 2
 
on specifies that intermittent error monitoring is to be used. To disable intermittent error monitoring, specify off. When you set this parameter to on, you can specify intermittent error conditions (the conditions used by the system to determine whether an intermittent error is occurring) by using the -intvl and -iemnum parameters. Specify the monitoring interval for an intermittent error in the -intvl parameter, and the number of times that the error is to occur in the -iemnum parameter. When these parameters are omitted and 3 or more errors occur within 30 minutes, the system assumes that an intermittent error is occurring.

(5) Setting the reservation level

You cannot set the reservation level (reservation control method for the LU).

(6) Setting the LU deletion functionality

You can select whether to enable the LU dynamic removal function.

For details on the LU deletion functionality, see 4.6.2 Deleting an LU dynamically.

Table 3-12 LU Deletion Function Values shows the values for the LU deletion functionality.

Table 3-12 LU Deletion Function Values

Values in the Options window Values in the set operation Description

OFF off The LU is not removed from HDLM-management even if the error occurs in all the paths to the LU, all the paths to the LU are disconnected, or the LU is deleted. The status of the paths is Offline(E) and Online(E). This operation is as equal as the operation in HDLM 05-01 or earlier. The off option is recommended when you want to use the same functionality as you use in HDLM 05-01 or earlier, without using the LU deletion functionality.

ON The Remove the LU even if there are Offline(C) paths option is set to OFF. on The LU is removed from HDLM-management when all the paths to the LU are disconnected. However, if an Offline(C) path is included in the disconnected paths, the LU is not deleted from HDLM-management.
The deleted LU is restored when it is recovered from the physical failure and the disk is re-scanned.

The Remove the LU even if there are Offline(C) paths option is set to ON. on -force The LU is removed from HDLM management when all the paths to the LU are disconnected, even when an Offline(C) path is included.
The removed LU is restored when it is recovered from the physical failure and the disk is re-scanned.
Setting by using the Options window
When using the Remove LU functionality, set it to ON. When not using the Remove LU functionality, set it to OFF. If you select ON, you can specify the conditions for LU dynamic removal in Remove the LU even if there are Offline(C) paths.
Setting the set operation
The following is an example to set the LU deletion functionality.
 
>dlnkmgr set -rmlu on -force
 
To enable the LU deletion functionality, specify on. To disable the LU deletion functionality, specify off. If you specify on, you can set the conditions for by using the -force parameter for LU dynamic removal.

Values in the Options window	Values in the set operation	Description
OFF	off	The LU is not removed from HDLM-management even if the error occurs in all the paths to the LU, all the paths to the LU are disconnected, or the LU is deleted. The status of the paths is Offline(E) and Online(E). This operation is as equal as the operation in HDLM 05-01 or earlier. The off option is recommended when you want to use the same functionality as you use in HDLM 05-01 or earlier, without using the LU deletion functionality.
ON	The Remove the LU even if there are Offline(C) paths option is set to OFF.	on	The LU is removed from HDLM-management when all the paths to the LU are disconnected. However, if an Offline(C) path is included in the disconnected paths, the LU is not deleted from HDLM-management. The deleted LU is restored when it is recovered from the physical failure and the disk is re-scanned.
The Remove the LU even if there are Offline(C) paths option is set to ON.	on -force	The LU is removed from HDLM management when all the paths to the LU are disconnected, even when an Offline(C) path is included. The removed LU is restored when it is recovered from the physical failure and the disk is re-scanned.

(7) Setting the error log collection level

The error log collection level can be set.

There are two error logs: the HDLM manager log file dlmmgrn.log (n indicates a file number from 1 to 16) and the HDLM GUI log file dlmguin.log (n indicates a file number of 1 or 2).

Table 3-13 Values for the error log collection level setting lists and describes values for the error log collection level setting.

Table 3-13 Values for the error log collection level setting

Value^# Description

0: Do not collect an error log No error logs are collected.

1: Error or higher level All information for errors with the level "Error" or higher is collected.

2: Warning or higher level All information for errors with the level "Warning" or higher is collected.

3: Information or higher level All information for errors with the level "Information" or higher is collected.

4: Information or higher level (including maintenance information) All information for errors with the level "Information" or higher (including maintenance information) is collected.

#

For the dlnkmgr command's set operation, specify a number such as 0 or 1.

If an error occurs, you might have to change the collection level to 1 or higher to collect log information.

The higher this value is set, the more log information will be output. When the output log is large, it takes less time to overwrite the old error log information with the new information.
Setting by using the Options window
Select Logging level from the pull-down list box.
Setting by using the set operation
The following shows an example of executing the command to set the error log collection level.
 
>dlnkmgr set -ellv 2
 
Specify the error log collection level in a number.

Value^#	Description
0: Do not collect an error log	No error logs are collected.
1: Error or higher level	All information for errors with the level "Error" or higher is collected.
2: Warning or higher level	All information for errors with the level "Warning" or higher is collected.
3: Information or higher level	All information for errors with the level "Information" or higher is collected.
4: Information or higher level (including maintenance information)	All information for errors with the level "Information" or higher (including maintenance information) is collected.

(8) Setting the trace level

The trace output level can be set.

A trace file in which a trace level can be set is hdlmtrn.log (n indicates a file number from 1 to 64).

Table 3-14 Values for the trace level setting lists and describes values for the trace level setting.

Table 3-14 Values for the trace level setting

Value^# Description

0: Do not output any trace No trace is output.

1: Error information only Only error information is output.

2: Summary of program operation Program operation summaries are output.

3: Details of program operation Program operation details are output.

4: All information All information is output.

#

For the dlnkmgr command's set operation, specify a number such as 0 or 1.

If an error occurs, you may have to set the trace level to 1 or higher to collect the log information.

The higher this value is set, the more log information will be output. When the output log is large, it takes less time to overwrite the old error log information with the new information.
Setting by using the Options window
Select Trace level from the pull-down list box.
Setting by using the set operation
The following is an example command to set the trace level.
 
>dlnkmgr set -systflv 1
 
Specify the trace level in a number.

Value^#	Description
0: Do not output any trace	No trace is output.
1: Error information only	Only error information is output.
2: Summary of program operation	Program operation summaries are output.
3: Details of program operation	Program operation details are output.
4: All information	All information is output.

(9) Setting the error log file size

The error log file size can be set.

There are two error logs: the HDLM manager log file dlmmgrn.log (n indicates a file number from 1 to 16) and the HDLM GUI log file dlmguin.log (n indicates a file number of 1 or 2).

You can specify a value (in kilobytes) from 100 to 2000000 for the error log file size. For HDLM GUI logs, the valid range of a file size is from 100 to 9900. If you specify a value of 9901 or more, 9900 is applied. The specified value is applied for HDLM manager logs.

When an error log file reaches the specified size, the information in the old error log file is replaced with new information, beginning with the oldest file. By specifying both the log file size and the number of log files, you can collect up to 32000000KB (approximately 30GB) of error logs in total.
Setting by using the Options window
For File size (for Error log), specify a value in kilobytes. To apply the error log file size you specified in the Options window of the HDLM GUI to the HDLM GUI log file dlmguin.log (n indicates a file number of 1 or 2), restart the HDLM GUI.
Setting by using the set operation
The following shows an example of executing the command to set the error log file size.
 
>dlnkmgr set -elfs 1000
 
Specify the size of the error log file in kilobytes.

(10) Setting the Number of Error Log Files

The number of the error log files can be set.

There are two error logs: the HDLM manager log file dlmmgrn.log (n indicates a file number from 1 to 16) and the HDLM GUI log file dlmguin.log (n indicates a file number of 1 or 2). The error logs you can set in the Options window and the set operation are limited to the HDLM manager logs file dlmmgrn.log (n indicates a file number from 1 to 16). The number of files for the HDLM GUI log file dlmguin.log (n indicates a file number of 1 or 2) is fixed to 2.

You can specify a value from 2 to 16 for the number of error log files (log files for the HDLM manager).

By specifying both the log file size and the number of log files, you can collect up to 32000000KB (approximately 30GB) of error logs in total.
Setting by using the Options window
For Number of files (for Error log), specify the number of error log files by using numeric characters.
Setting by using the set operation
The following shows an example of setting the number of error log files.
 
>dlnkmgr set -elfn 5
 
Specify the number of error log files by using numeric characters.

(11) Setting the trace file size

The trace file size can be set.

Trace files for which a trace file size can be set are hdlmtrn.log (n indicates a file number from 1 to 64). The length of a trace file is fixed. Thus, even if the amount of written trace information is less than the set file size, the file size of each output trace file is always fixed.

For the trace file size, you can specify a value (in kilobytes) from 100 to 16000. If you specify a value smaller than the setting value, the message (KAPL02080-W for the Options window and KAPL01097-W for the set operation) is displayed to confirm execution, and the trace file is temporarily deleted.

Once trace data has been written to all the trace files, the oldest file is overwritten with new trace data.

By specifying both the trace file size and the number of trace files, you can collect up to 1024000KB of trace data in total.
Setting by using the Options window
For File size (for Trace), specify a value in kilobytes.
Setting by using the set operation
The following shows an example of setting the trace file size:
 
>dlnkmgr set -systfs 2000
 
Specify the size of the trace file in kilobytes.

(12) Setting the number of trace files

You can set the number of trace files.

Trace files for which the number of files can be set are hdlmtrn.log (n indicates a file number from 1 to 64).

For the number of trace files, you can specify a value from 2 to 64. If you specify a value smaller than the setting value, the message (KAPL02080-W for the Options window and KAPL01097-W for the set operation) is displayed to confirm execution, and the trace file is temporarily deleted.

By specifying both the trace file size and the number of trace files, you can collect up to 1024000KB of trace data in total.
Setting by using the Options window
For Number of files (for Trace), specify the number of trace files by using numeric characters.
Setting by using the set operation
The following shows an example of setting the number of trace files:
 
>dlnkmgr set -systfn 10
 
Specify the number of trace files by using numeric characters.

[Contents][Back][Next]

[Trademarks]