Re: Clustering Support
Posted in 1993
Topics: Clustering, Grid & MACH11
In <1lv7nqINN997@emory.mathcs.emory.edu> ckipp@vth1.vth.colostate.edu (Cathy Kipp) writes: >Mark, >Yes, Informix does support clustering and on all platforms as far I can tell. >Use the create index command with the cluster feature. This will create a new >index with clustering. You may only use one clustered index on a table. The >clustering will not be updated as changes are made to the table, so you will >periodically need to drop and recreate the clustered index. >Cathy Thanks for your reply Cathy. Unfortunately, my question was ambiguous, so let me be more specific. By "clustering," I mean the ability for multiple copies of a database kernel running on multiple nodes in a network to each share access to a single, central database. For example, suppose I have four Unix boxes in a network, and all four boxes share access to a set of disks on which one Oracle database resides. With Oracle Parallel Server, I can start an Oracle kernel on each node in the network, and have each open and mount the shared database. I can then run transactions against that database from anywhere in the network, and Oracle will automatically take care of the locking (down to the database block level) required to make such concurrent access work. In addition, if a software or hardware problem prevents a kernel on a given node from running, I can just switch to one of the remaining nodes and keep right on working. Does Informix support this? Thanks, Mark -- Mark A. Peters ****** ====================== Control Data Systems, Inc. ****** == "What a save!!!" == Internet: mark.a.peters@cdc.com ****** == "What an idea!!" ==