Home | Previous Page | Next Page   Configuration > Coserver Configuration > Managing Coservers >

Using the Default Cogroups

In earlier versions of Extended Parallel Server, one cogroup was created automatically and managed by the database server, cogroup_all. To distinguish among the three categories of coservers, however, cogroup_all, which still exists, is divided into three new groups, one cogroup for each coserver category, as follows:

When you add a specific-purpose coserver, as described in Adding Coservers, the coserver is automatically added to appropriate default cogroups. These cogroups are determined by the capability you assign to the coserver. For example, if you specify COMPUTE capability for a coserver, that coserver belongs to cogroup_compdata and cogroup_all. If you specify both COMPUTE and TEMPDATA capabilities, the coserver belongs to both cogroup_tempdata, cogroup_compdata, and cogroup_all.

Tip:
If you do not plan to use the specific-purpose coservers, you do not need to stop using cogroup_all in dbslice creation scripts. However, to prevent future conflicts between the type of dbspace on the host coserver and the capabilities of the coserver, use the new cogroup_permdata cogroup.
Home | [ Top of Page | Previous Page | Next Page | Contents | Index ]