Job scheduler (jsh) and DistribuSys

Records in the Jsh (job scheduler) table schedule jobs for execution. In a DistribuSys environment, each Jsh record has a dataset.

The value of this dataset determines where the record is replicated, as well as the ownership privileges associated with the record. The ownership privilege controls the systems that are permitted to execute the job. The dataset also defines the systems that should receive record configuration changes.

The jsh table treats dataset 0 as a local dataset. The functionality allows you to locally execute a jsh record with a dataset value of 0, rather than outside of the local services by the RealTime replication server. All of the other tables treat dataset 0 as an incomplete record definition, and do not process the record.

When a system is set up initially, all of the jsh records have their dataset values set to 0, and all of the systems get identical copies of these jsh records. Systems are able to modify their jsh records to suit their own needs, and those changes are not replicated to other systems. This means, for example, that only one record is required for the jsh savedata job. Instead of having one record for each system, this record can be modified on each system as needed. The advantage is that fewer records are required.

Records for the current system are processed by the job scheduler if one of the following applies:

• The jsh record has a dataset value of 0.

• The privileges that are associated with the jsh record non-zero dataset indicate that it is owned by the current system.

In cases with non-zero datasets, jsh acts like all of the other tables as regards to execution and replication. The jsh record executes only in the owning system and replicates changes only to systems that belong to that dataset privilege.