Re: 21 terabytes at NYNEX

David M Walker (davidw_at_datamgmt.com) wrote:
: Building a system to hold 21Tb has its problems. These are not
: really related to the file system size of 2Gb as the database
: would ahve to be raw anyway.
 

: If we assume 10Gb per disk, 12 disks per SCSI channel, 4 channels
: per Disk controller card you would need 44 disk controllers.
 

: In practice most SMP platforms support about 16 mid plane slots
: (Though look at some of the new offerings from Sequent and others)
 

: If we need 8 slots for processors and 2 slots for memory then we
: are left with six slots for controllers. This gives us about 2.9Tb
: of raw disk to start with. In practice, this would reduce down to
: 2Tb after RAID, O/S, Swap and load areas where taken into consideration.

Unfortunately you are not thinking about the capabilities of the Digital platform. Using CI (cluster interconnect) you could put x CI controllers (at 70-100Mbits/sec) that can talk to 3-400 HSx Storage controllers which can handle a very large number of disk each.

: 21Tb therefore needs to be served by 12 Systems with a distributed
: system, or more likely a using a series of data marts built on
: a transaction repository.

If you want REAL scalability use OracleRDB and VMS. It is a lot more stable and 1000 times more manageable when you get into database of this size. Especially since your filesystems can only be 2GB in Unix. VMS does not have this restriction. RDB (V7.0) allows for both horizontal AND vertical partitioning of your data. Does Oracle? And you would want to use this to manage this size of database. There are projects using this configuration that will exceed 100Tb in the next few years.

: I wrote an article on this scale of system for the European Oracle
: User Group and have built systems upto 5Tb.

Michael Austin
OracleRDB Independent Consultant
austin_at_inmind.com Received on Tue May 07 1996 - 00:00:00 CEST