DOEACC Society 2006 DOEACC C Level C12 Distributed Systems ( ) - Question Paper

the outcomes.
C12-R3 Page two of three January, 2006
b) Why do DNS root servers hold entries for two-level names such as yahoo.com and
engineering.edu, rather than one-level names such as edu and com?
c) GNS does not guarantee that all copies of entries in the naming database are up-to-date.
How are clients of GNS likely to become aware that they have been provided an out-of-date
entry? Under what circumstances might it be harmful?
(6+6+6)
4.
a) Which features of the AFS design make it more scalable than NFS? elaborate the limits on
its scalability, assuming that servers can be added as required? Which latest
developments offer greater scalability?
b) Draw and discuss the typical infrastructure components for a simple multimedia
conferencing application running on 2 personal computers.
c) In the subsequent table a few typical kinds of multimedia applications are provided. Fill up the
approximate data rates and frame or sample size and frequencies.
Data Rate Sample or Frame
(approximate) Size Frequency
i) Telephone speech
ii) CD-quality sound
iii) Standard TV video (uncompressed)
iv) Standard TV video (MPEG-1
compressed)
v) HDTV video (uncompressed)
vi) HDTV video (MPEG-2 compressed)
(8+6+4)
5.
a) Estimate the time needed to crack a 56-bit DES key by a brute-force attack using a 500
MIPS (million instruction per second) workstation, assuming that the inner loop for a bruteforce
attack program involves around 10 instructions per key value, plus the time to encrypt
an 8-byte plaintext. Perform the identical computation for a 128bit IDEA key. Extrapolate your
computations to find the cracking time for a 50,000 MIPS parallel processor (or an Internet
consortium with similar processing power).
b) How would email be sent to a large list of recipients using PGP (Pretty Good Privacy) or a
similar scheme? Suggest a scheme that is simpler and faster when the list is used
frequently.
(10+8)
6.
a) discuss why serial equivalence requires that once a transaction has released a lock on an
object, it is not allowed to find any more locks. A server manages the objects a1, a2, … an.
The server provides 2 operations for its clients:
learn (i) returns the value of ai
write (i, Value) assigns Value to ai
The transactions T and U are described as follows:
T: x= learn (i); write (j, 44);
U: write (i, 55); write (j, 66);
define an interleaving of the transactions T and U in which locks are released early with
C12-R3 Page three of three January, 2006
the effect that the interleaving is not serially equivalent.
b) The role of server and its status at the time of failure are provided beneath. Write and explain the
action of the recovery manager with respect to the two-phase commit protocol for any
transaction.
Server Role Status Action of recovery manager
i) Coordinator prepared
ii) Coordinator committed
iii) Participant committed
iv) Participant uncertain
v) Participant prepared
vi) Coordinator done
(9+9)
7.
a) A file server uses caching, and achieves a hit rate of 80%. File operations in the server cost
5ms of CPU time when the server obtains the requested block in the cache, and take an
additional 15ms of disk I/O time otherwise. Explaining any assumptions you make, estimate
the server’s throughput capacity (average requests/sec) if it is:
i) Single-threaded;
ii) Two-threaded, running on a single processor;
iii) Two-threaded, running on a two-processor computer.
b) Prove that Ivy’s write-invalidate protocol guarantees sequential consistency.
c) In Ivy’s dynamic distributed manager algorithm, what steps are taken to minimize the
number of lookups necessary to obtain a page?
(6+6+6)
C12-R3 Page four of three January, 2006

Earning:   Approval pending.