When using OpenTP1 in a cluster system or parallel-processing system, OpenTP1 provides the Global Archive Journal facility to simplify the unloading of system journal files from multiple nodes. Rather than having to unload the standby system journal file for each individual node, a user can:

Use the OpenTP1 Global Archive Journal facility to automatically copy the standby system journal files from a number of nodes into an archive journal file on a single dedicated node.

(For details on the Global Archive Journal facility, see 6.2.3 Global Archive Journal facility.)

1. Use the command jnlunlfg to unload the contents of the archive journal file to a file called the global archive unloaded-journals file.

For global archive journals, you can select the kind of the journal to be archived. Specify the kind of the journal record by the jnl_arc_rec_kind operand in the system journal definition. For details of the kinds of journals, see 4.2.2 System files: system journal files.

Up to 16 types of archive journal files can be created. The created archive journal files are within the archive-journal target node, which is the node used for archiving. As with system journal files, individual archive journal files are logical filegroups containing physical files.

Before using an archive journal file, you must specify whether the file is to be opened when OpenTP1 starts. By specifying the filegroup name in a command parameter (e.g., in jnlopnfg or jnlls) an archive journal file can be operated on in logical filegroup units.

Archive journal files can be grouped into resource groups. A resource group is a group of archive journal files that are separated by intended use. The global archive journal service uses the names of resource groups to manage archive journal files. Up to 20 nodes can be archived in one resource group.

To create an archive journal file, create a physical file in the OpenTP1 file system using the jnlinit command, and specify the archive journal file in the archive journal service definition.

In the archive journal service definition where the resource group is to be defined, specify the filegroup name, physical file name, and element file name.

• Physical file name
  Pathname that indicates the real OpenTP1 file.
  
• Element file name
  A name given to the element file when using the parallel access facility for archive journal files.
  

An element file can be duplicated. A filegroup name is a name given to a group of one or more element files. When not using the parallel access facility for archive journal files (when there is only one element file in the filegroup), the element file name can be omitted. Up to 8 element files can be contained in a filegroup.

Before using an archive journal file, specify, for each filegroup, whether to open the archive journal file at the same time OpenTP1 starts. By specifying a filegroup name in a command argument, the archive journal file can be operated in logical filegroup units. Note that one resource group requires two or more filegroups.

An archive journal file can be used by configuring a filegroup using multiple element files. This facility is called the parallel access facility for archive journal files.

In an archive journal file, the journals for multiple nodes are constantly input and output. Sometimes the input and output operations for the archive journal file cannot keep up with the journals that need to be archived. In this situation, configure a filegroup using more than one element file to access the archive journal file in parallel. This increases the performance of the archive.

Physical files that make up a filegroup can use the same SCSI interface and hard disk; however, when they do, the SCSI interface and the hard disk cannot be accessed in parallel, nor can they provide the full parallel access facility. Try to use different SCSI interfaces and hard disks. The sizes of physical files need not be the same. If the sizes are different, however, OpenTP1 considers the smaller file to be the size of the filegroup. To use resources effectively, match the sizes of physical files whenever possible.

When using the parallel access facility for archive journal files, the number of element files that can access in parallel (number of parallel accesses) decreases if an error occurs in the archive journal file. To make sure the input and output operations of the archive journal file keep up with the journals that need to be archived, you can specify the minimum number of parallel accesses to be guaranteed. This minimum number of parallel accesses to be guaranteed is the minimum number of distributions. The maximum number of distributions is the maximum number of parallel accesses.

When specifying a filegroup in the archive journal service definition, use the element file name. Specify the maximum number of distributions in the jnl_max_file_dispersion operand and the minimum number of distributions in the jnl_min_file_dispersion operand in the archive journal service definition.

An archive journal file can be duplicated by configuring an element file using two physical files. If a filegroup is operated with the duplicated archive journal file, OpenTP1 outputs the same journal in system A and in system B.

If a failure occurs in one system while reading the checkpoint dump, the same data can be read from the other system. In this way, duplicating a checkpoint dump file increases reliability.

(a) Specifying the archive journal file definition

To duplicate an archive journal file, specify Y in the jnl_dual operand in the archive journal file definition. At this time, specify two physical files (system A and system B) in one filegroup. You should store the physical file of system A and the physical file of system B on separate disks to prevent a failure in both physical files at the same time. The size of the physical file of system A and that of system B need not be the same. If the sizes are different, however, OpenTP1 considers the smaller file to be the size of the journal. To use resources efficiently, match the sizes whenever possible.

(b) Single-system operation and non-single-system operation

When an archive journal file is duplicated, single-system operation or non-single-system operation can be selected.

• Single-system operation
  If either of the two physical files in an element file can be used, the element file can be used.
  
• Non-single-system operation
  The element file can be used only when both physical files in the element file can be used.
  

The archive journal service definition allows or disallows single-system operation. When single-system operation is permitted, even if an error occurs in one physical file, the system continues operating with the remaining physical file until the error is corrected. During single-system operation, the archive journal file is not duplicated, decreasing reliability.

An archive journal file can have two statuses, the same as a system journal file.

• Available
  The number of usable element files among the element files in a filegroup is greater than the required number.
  
• Unavailable
  The number of usable element files among the element files in a filegroup is less than the required number.
  

An available element file implies that the required number of physical files is opened in that element file. An element file is unavailable if the required number of physical files are not opened in the element file. The required number of element files and the required number of physical files are determined according to the values specified in the archive journal service definition.

Filegroups are managed as having current or standby status if available for journal output, or in reserved status if unavailable for journal output. The required number of element files and physical files are determined by the values specified in the archive journal service definition.

• Current
  An available filegroup which is the current destination of journals to be output. There is always only one current filegroup.
  
• Standby
  An available filegroup which is not currently the destination of journals to be output but is waiting to be switched to become current. The two standby statuses are:
  
  • Ready to be the next swap destination:
    A standby filegroup that can be overwritten (there is no journal that is required for recovery) and is already unloaded.
    
  • Not ready to be the next swap destination:
    A standby filegroup that cannot be overwritten or is waiting to be unloaded.
    
• Reserved
  An unavailable filegroup.
  

An archive journal file requires at least two filegroups other than the reserved filegroups.

When an archive journal file is waiting for an unload, you should use the command jnlunlfg to unload. A filegroup of an archive journal file which has not been unloaded might cause a complete lack of swappable files, which would cause an abnormal termination of the global archive journal service.

As with a system journal file, an unloaded ordinary file allows DAM file recovery and editing of files containing operation statistics. Also, a user journal can be inherited by an offline program.

A global archive unloaded-journals file contains system journal information about multiple OpenTP1 nodes. You can edit this journal: for example, by editing merged operation statistics or by sorting multiple OpenTP1 node journals by time.

(a) Restraining an unload check

The standby filegroup which is usually in the unload wait status can be specified to use as it is. This is called a unload check restraint. When restraining an unload check, specify N to the jnl_unload_check operand in the system journal service definition. To restrain an unload check, specify N in the jnl_unload_check operand in the archive journal service definition.

If the unload check is restrained, do not unload the journal files using a command during the operation of OpenTP1. If an unload is performed by specifying an unload check restraint, unload the file after the files are once closed by the jnlclsfg command.

When the unload check is restrained, the journal edit commands (jnlcolc command, jnlstts command) cannot be executed as input for a global archive unload journal file because the global archive unload journal file cannot be created.

If an unload check is restrained, the unload wait status may be output when confirming the filegroup status using the jnlls command. Even if the unload wait status is output, the current filegroups are actually used.

In the system journal services related to the Global Archive Journal facility, you can add the following journal file statuses: archived or not archived. These indicate whether or not a system journal filegroup has been archived. You can display this status of a system journal file with the jnlls command.

After a system journal filegroup is copied to an archive journal file, the system journal filegroup is eligible to be swapped even if the archive journal file has not yet been unloaded.

Some standby archive journal filegroups can become swap destinations while others cannot. Those that can become swap destinations are filegroups that are already unloaded. Filegroups that are not unloaded cannot be swap destinations. If there is no swap destination for the archive journal file, OpenTP1 terminates abnormally: see 5.3.3(1)(d) Recovering from system-file errors: archive journal files.

Figure 4-12 shows the swapping of archive journal files.

Figure 4-12 Swapping archive journal files