23.2 Estimating the database area
The following describes an estimate of the database area, which is required when you want to customize the database area.
In JP1/AJS3, a database can be created as a large-scale, medium-scale, or small-scale database. When JP1/AJS3 is installed as a new installation, an embedded database is set up as a small-scale database with the database auto-increment function already configured. Use the following values as guidelines for the disk space requirements of the small-scale model.
-
Total number of units: 5,000 (If you execute the ajsembdbsetup command with the -e sjis option specified to set up the database for the scheduler service, 4,000)
This includes 4,000 jobs (PC jobs, Unix jobs, flexible jobs, and QUEUE jobs). (If you execute the ajsembdbsetup command with the -e sjis option specified to set up the database for the scheduler service, 3,200)
-
Number of logs to keep: 5 (If you execute the ajsembdbsetup command with the -e sjis option specified to set up the database for the scheduler service, 4)
You can also set the database scale by executing the jajs_setup or jajs_setup_cluster command with the -M option or by performing an advanced setup on an embedded database. You can change the scale of a database by performing an advanced setup of the embedded database. To change the scale, you need to remove the setup of the embedded database, and then perform an advanced setup of the embedded database.
Although JP1/AJS3 allows you to set up and use multiple embedded database environments, the following describes the estimation method for a single embedded database environment. If you want to set up multiple scheduler services in a single embedded database environment, add the estimates for all scheduler services in the environment.
The following table describes the database models that can be specified during setup of the embedded-database environment. Make sure that the disk has extra free space, because some file systems require more disk space than described in the table.
The following estimate assumes that you perform regular database maintenance. If you are using the database auto-increment function, the actual size might be bigger than the estimated size unless you perform regular database maintenance.
|
Scale |
Purpose |
|---|---|
|
Large |
Total number of units: About 48,000 to 240,000 Number of units executed per day: 30,000 to 120,000 Disk space guideline: About 20,700 MB Number of logs to keep: 5 (data area: about 6,700 MB, system area: about 14,000 MB) When the database auto-increment function is used: If the total number of estimated units is larger than expected, add the following to the disk space guideline: About 20,700 + 0.1 x (total number of units - 240,000) MB |
|
Medium |
Total number of units: About 5,000 to 48,000 Number of units executed per day: 5,000 to 30,000 Disk space guideline: About 4,200 MB Number of logs to keep: 5 (data area: about 1,400 MB, system area: about 2,800 MB) When the database auto-increment function is used: If the total number of estimated units is larger than expected, add the following to the disk space guideline: About 4,200 + 0.1 x (total number of units - 48,000) MB |
|
Small |
Total number of units: About 5,000 or fewer Number of units executed per day: 5,000 or fewer Disk space guideline: About 520 MB Number of logs to keep: 5 (data area: about 200 MB, system area: about 320 MB) When the database auto-increment function is used: If the total number of estimated units is larger than expected, add the following to the disk space guideline: About 520 + 0.1 x (total number of units - 5,000) MB |
For example, if the total number of units falls into the large scale range (48,000 to 240,000), the number of units executed per day is in the medium scale range (5,000 to 30,000), and the operation does not meet the usage in the above table, calculate as follows:
(0.0045# x a) + (0.001 x b) + (c x 0.0065) (unit: MB)
- Legend:
-
- a:
-
Substitutes for the total number of units.
- b:
-
Substitutes for the total number of waits, which is defined for every unit with wait conditions.
- c:
-
Substitutes for the value calculated by using the expression below for all root jobnets to be registered by execution registration.
Total-number-of-root-jobnets-and-subsequent-units x (number-of-generations-to-be-saved + d)
If LEGACY is specified in the SAVEGENTYPE environment parameter, use the following calculation expression for the root jobnets with startup conditions set.
Total-number-of-root-jobnets-and-subsequent-units x (number-of-generations-to-be-saved + d + number-of-generations-to-be-saved x number-of-generations-to-be-saved)
- d:
-
Substitute a following value, according to the execution registration method.
Immediate execution registration: 1
Planned execution registration: 1
Fixed execution registration: number of scheduled generations
- #
-
The calculation is made assuming the total of the values entered in the unit definition, such as the command statement, executable file name, and script file name, is 2 KB per unit.
When using the jobnet release function, calculate the size of the root jobnet in the release source unit that is set for release and the units contained in that jobnet, and then add that value.
If the calculated value is larger than 200 MB, we recommend that you configure the embedded database environment using a medium-scale model. If the value is larger than 1,400 MB, we recommend that you configure the embedded database environment using a large-scale model.
If the calculated value is larger than 6,700 MB, we recommend that you configure the embedded database environment using a large-scale model, and extend the data area after configuration. For details about the data area extension, see ajsembdbaddarea in 3. Commands Used for Normal Operations in the manual JP1/Automatic Job Management System 3 Command Reference. Note that if you use the ajsembdbaddarea command to extend the data area, extend the table area and the index area at a ratio of 5:2.
When you estimate the size of the system area (the area for system log files), estimate the size of the system area for the scale determined when the data area size was estimated. If the required system area exceeds the system area for the large-scale model, use the following expressions for estimation. If there is no area larger than that estimated, expand the system log files.
-
Unload log operation: Data-area x 2 (in megabytes)
-
Operation without using system logs: (0.0045# x a) x 2 (in megabytes)
- Legend:
-
a: Total number of units
- #
-
The calculation is made assuming the total of the values entered in the unit definition, such as the command statement, execution file name, and script file name, is 2 KB per unit.
For details about the system log file extension, see ajsembdbaddlog in 3. Commands Used for Normal Operations in the manual JP1/Automatic Job Management System 3 Command Reference.
The following is an example estimate for the database model in an embedded database environment.
- Example:
-
The following case serves as an example:
The environment setting parameter SAVEGENTYPE: TOTAL
Total number of units: 200,000
Total number of waits defined in all units with wait conditions: 1,000
Number of units executed per day: 10,000
Number of generations to be saved: 5
Number of scheduled generations that registered a jobnet for fixed execution: 10
(0.0045 x 200,000) + (0.001 x 1,000) + (10,000 x (5 + 10) x 0.0065)
= 900 + 1 + 975
= 1,876 (MB)
In this case, we recommend that you configure the embedded database environment using a large-scale model.
If you execute the ajsembdbsetup command with the -e sjis option specified to set up the database of the scheduler service, the database models are as shown in the following table.
|
Scale |
Purpose |
|---|---|
|
Large |
Total number of units: About 38,400 to 192,000 Number of units executed per day: 24,000 to 96,000 Disk space guideline: About 20,700MB Number of logs to keep: 4 (data area: about 6,700 MB, system area: about 14,000 MB) When the database auto-increment function is used: If the total number of estimated units is larger than expected, add the following to the disk space guideline: About 20,700 + 0.125 x (total number of units - 192,000) MB |
|
Medium |
Total number of units: About 4,000 to 38,400 Number of units executed per day: 4,000 to 24,000 Disk space guideline: About 4,200 MB Number of logs to keep: 4 (data area: about 1,400 MB, system area: about 2,800 MB) When the database auto-increment function is used: If the total number of estimated units is larger than expected, add the following to the disk space guideline: About 4,200 + 0.125 x (total number of units - 38,400) MB |
|
Small |
Total number of units: About 4,000 or fewer Number of units executed per day: 4,000 or fewer Disk space guideline: About 520 MB Number of logs to keep: 4 (data area: about 200 MB, system area: about 320 MB) When the database auto-increment function is used: If the total number of estimated units is larger than expected, add the following to the disk space guideline: About 520 + 0.125 x (total number of units - 4,000) MB |
For example, if the total number of units falls into the large scale range (38,400 to 192,000), the number of units executed per day is in the medium scale range (4,000 to 24,000), and the operation does not meet the usage in the above table, calculate as follows:
((0.0045# x a) + (0.001 x b) + (c x 0.0065)) x 1.25 (MB)
- Legend:
-
- a:
-
Substitutes for the total number of units.
- b:
-
Substitutes for the total number of waits, which is defined for every unit with wait conditions.
- c:
-
Substitutes for the value calculated by using the expression below for all root jobnets to be registered by execution registration.
Total-number-of-root-jobnets-and-subsequent-units x (number-of-generations-to-be-saved + d)
If LEGACY is specified in the SAVEGENTYPE environment parameter, use the following calculation expression for the root jobnets with startup conditions set.
Total-number-of-root-jobnets-and-subsequent-units x (number-of-generations-to-be-saved + d + number-of-generations-to-be-saved x number-of-generations-to-be-saved)
- d:
-
Substitute a following value, according to the execution registration method.
Immediate execution registration: 1
Planned execution registration: 1
Fixed execution registration: number of scheduled generations
- #
-
The calculation is made assuming the total of the values entered in the unit definition, such as the command statement, executable file name, and script file name, is 2 KB per unit.
When using the jobnet release function, calculate the size of the root jobnet in the release source unit that is set for release and the units contained in that jobnet, and then add that value.
If the calculated value is larger than 200 MB, we recommend that you configure the embedded database environment using a medium-scale model. If the value is larger than 1,400 MB, we recommend that you configure the embedded database environment using a large-scale model.
If the calculated value is larger than 6,700 MB, we recommend that you configure the embedded database environment using a large-scale model, and extend the data area after configuration. For details about the data area extension, see ajsembdbaddarea in 3. Commands Used for Normal Operations in the manual JP1/Automatic Job Management System 3 Command Reference. Note that when you use the ajsembdbaddarea command to extend the data area, extend the table area and the index area at a ratio of 5:2.
Estimate the system area (system log file) by using the following formula. If the space is insufficient, extend the system log file.
When you estimate the size of the system area (the area for system log files), estimate the size of the system area for the scale determined when the data area size was estimated. If the required system area exceeds the system area for the large-scale model, use the following expressions for estimation. If there is no area larger than that estimated, increase the number or size of the system log files.
-
Unload log operation: Data-area x 2 (in megabytes)
-
Operation without using system logs: (0.0045# x a) x 1.25 x 2 (in megabytes)
- Legend:
-
a: Total number of units
- #
-
The calculation is made assuming that the total of the values entered in the unit definition, such as the command statement, execution file name, and script file name, is 2 KB per unit.
For details about the system log file extension, see ajsembdbaddlog in 3. Commands Used for Normal Operations in the manual JP1/Automatic Job Management System 3 Command Reference.
The following is an example estimate for the database model in an embedded database environment.
- Example:
-
The following case serves as an example:
The environment setting parameter SAVEGENTYPE: TOTAL
Total number of units: 100,000
Total number of waits defined in all units with wait conditions: 1,000
Number of units executed per day: 10,000
Number of generations to be saved: 5
Number of scheduled generations that registered a jobnet for fixed execution: 10
((0.0045 x 100,000) + (0.001 x 1,000) + (10,000 x (5 + 10) x 0.0065)) x 1.25
= (450 + 1 + 975) x 1.25
= 1782.5 (MB)
In this case, we recommend that you configure the embedded database environment using a large-scale model.
If the system area for the embedded database is duplexed during setup of the embedded-database environment, more disk space is required. The following table shows how much additional disk space is required.
|
Scale |
Amount of additional disk space required |
|---|---|
|
Large |
About 14,000 MB |
|
Medium |
About 2,800 MB |
|
Small |
About 320 MB |
You can create the data area and system area for the embedded database in RAW files. For this purpose, the following two or three partitions are required:
-
Data area partition (ajssys01)
-
System area partition 1 (ajssys11)
-
System area partition 2 (ajssys17)#
The following table shows the partition sizes required for each scale.
|
Scale |
Purpose |
|---|---|
|
Large |
ajssys01: About 6,700 MB ajssys11: About 14,000 MB ajssys17: About 14,000 MB# |
|
Medium |
ajssys01: About 1,400 MB ajssys11: About 2,800 MB ajssys17: About 2,800 MB# |
|
Small |
ajssys01: About 200 MB ajssys11: About 320 MB ajssys17: About 320 MB# |
- #:
-
This partition is required only if the system area for the embedded database is duplexed.
- About the database auto-increment function:
-
If you use the database auto-increment function, the amount of allocated data area is automatically increased a little bit each time it becomes insufficient. The function can expand the data area to the maximum extent possible on the disk on which the data area is located. The files that can be expanded by this function are as follows:
-
The ajssys041 and ajssys042 files in the data area storage directory
-
The files created under the directory specified for the -d option in the ajsembdbaddarea command
Similarly, if you use the system log auto-increment function, the amount of allocated system log file is also automatically increased a little bit each time it becomes insufficient. The system area can be expanded to three times the initial size. Note that the initial size of a system log file is 23 megabytes for small scale models, 224 megabytes for medium scale models, and 1,152 megabytes for large scale models.
The system log files that can be expanded by this function are as follows:
-
The ajssys01101 to ajssys01112# files in the data area storage directory
-
The files created under the directory specified for the -d option in the ajsembdbaddlog command, for which sys is specified for the -r option
- #:
-
If system file area creation directories 1 and 2 are specified in the ajsembdbbuild command during advanced setup of an embedded database, the following files are expanded:
-
The ajssys01101 to ajssys01112 files in system file area creation directory 1
-
The ajssys01701 to ajssys01712 files in system file area creation directory 2
-
For details about the ajsembdbaddarea and ajsembdbaddlog commands, see 3. Commands Used for Normal Operations in the manual JP1/Automatic Job Management System 3 Command Reference.
-
- Supplementary notes
-
-
For details about the required disk space when system files are duplicated, see Table 23-2 Disk space required for the embedded database operating environment in 23.1.1 Examining the embedded-database operating environment and operating method.
-
When advanced setup of the embedded database is used to create an embedded database that will be used for operation that uses unloading, you need to estimate the disk space for storing unload log files. For details about operation that uses unloading, see F.1(1) Restoring the scheduler database from unload log files in the JP1/Automatic Job Management System 3 System Design (Configuration) Guide.
-
For details about how to estimate the disk space after migration from a version 8 embedded database, see 8.5.2(7) Estimating the disk space requirements after migration from a version 8 embedded database in the JP1/Automatic Job Management System 3 System Design (Configuration) Guide.
-