robinson@ecsvax.UUCP (Gerard Robinson) (12/29/86)
[] The Office of Information Systems of the UNC School of Medicine is charged with creating a computing system which will connect the business offices of the School's various departments and Centers for sharing a database of administrative and clinical data. Administrative data consists of University-level information on funding sources and personnel, as well as School-level information such as sub-departmental cost centers and common appointment and meeting calendars. Clinical data will ultimately comprise patient histories, appointments, and billing information as well as such data as case libraries for instructional purposes. At issue is the hardware and software environment for this system. Our current operating environment includes a VAX-11/780 running VMS and Relational Technology's INGRES using only asynchronous terminals or PCs as terminals for distrubuted access. All or nearly all of the business offices have IBM compatible personal computers (I'm sure there are Macintoshs in some) and the University supports a broadband communications system to which all of the School's buildings (>15 on 1/2 sq. mi.) have access. Few of the business office PCs are connected to a high-performance (>1Mbps) network, and the currently supoorted broadband data services are only 19.2kbps or slower asynchronous and 128kbps or less point-to-point synchronous communications with Bridge Communications IEEE 802.3 broadband-ethernet gateways on order. Our particular interest is in evolving to a distributed computing system which has a 5-10Mbps network as its communications backbone and uses (but not stretches) the currently available equipment in connection with more advanced (but still relatively proven) technologies. For us, this description brings immediately to mind a Sun Microsystems distributed workstation environment. We are aware of potential substitutes for the Suns in the form of the oft rumored VAXstar, desktop VAX, as well as some other vendors equipment, and would like to hear about your favorites. Information regarding current communications technology for broadband networks would also be helpful. Inexpensive means of connecting CPU's directly to a higher-performance broadband network would be especially exciting and unexpected. Reliability is also a key issue, since the systems will come to be relied upon by physicians for patient care information. Information regarding database systems whose performance or ease of programmability relative to RTI's INGRES would warrant our conversion effort is of great interest as well. Any feedback will be most appreciated. Thanks for reading this far, and thanks in advance for your responses. Gerard A. Robinson ...decvax!mcnc!ecsvax!robinson (uucp) UGAR@TUCC.BITNET