Post Reply 
(SR-52) Queuing Network Models
08-17-2020, 12:29 PM
Post: #1
(SR-52) Queuing Network Models
An excerpt from Some Queuing Network Models of Computer Systems, E. S. Herndon, NASA/JSC, M80-7, JSC 16494, MARCH 1980, 56 pgs.

"          ABSTRACT
  Queuing Network Models of a computer system operating with a single workload type are presented.  Programs which operate on the Texas Instruments SR-52 programmable calculator are included …

  In an earlier paper … conventional Markov modeling techniques were used to develop a simple model of n terminals dealing with a single server system. A program for the Texas Instruments SR-52 programmable calculator was presented in that paper … Four programs are presented in this paper. They provide a capability to handle a large number of closed network, single workload problems …

  Four programs have been developed to aid in the analysis of closed queuing networks.
    1. Batch model with homogeneous service times
    2. Batch model with one load dependent server
    3. Interactive model with up to five devices
    4. Interactive model with a load dependent central server
  The two programs for batch models will be discussed together since there are only minor variations between the two. Then the interactive models will be presented …

… paper provides listings of the four programs and a sample case study which can be replicated on the SR-52


ps: download link is 19800017584.pdf in bottom left corner
Find all posts by this user
Quote this message in a reply
08-18-2020, 11:08 AM
Post: #2
RE: (SR-52) Queuing Network Models
I would like everyone to know the theories behind these calculations. For more than 20 years I have struggled to explain how to smooth peak loads while ensuring a maximum level SLA.
This is especially important for designing services APIs that do not all have the same response time.

Thibault - not collector but in love with the few HP models I own - Also musician :
Find all posts by this user
Quote this message in a reply
Post Reply 

User(s) browsing this thread: