1. Chien-Chung Shen CIS/UD cshen@udel.edu
Review of Final Exam
Chien-Chung Shen
CIS/UD

2. Semaphore Usage: Summary
Mutual exclusion: binary semaphore as mutex lock
Controlled access to a given resource consisting of a finite number of instances: counting semaphore
semaphore is initialized to the number of instances available
Synchronization:
two concurrent running threads T1 and T2 with statements S1 and S2, respectively
require S2 be executed only after S1 has completed (on one CPU)
Semaphore s = 0;
T1: S1; signal(s);
T2: wait(s); S2; S1 S2 T1 T2

3. Bounded Buffer Multiple P/C

4. Buddy Allocation
search for free space recursively divides free space by two until a block that is big enough to accommodate the request is found
(William Stallings: OS)

5. $ ln Chapter3 Chapter3.hard $ ls –il (show attributes of files)
two names for the same file
$ ln Chapter3 Chapter3.hard
$ ls –il
(show attributes of files)

6. Get Info about Files
[cisc361:/usa/cshen/ ] echo hello > foo
[cisc361:/usa/cshen/ ] more foo
hello
[cisc361:/usa/cshen/ ] stat foo
  File: 'foo'
  Size: 6         Blocks: 3          IO Block: regular file
Device: 29h/41d Inode:       Links: 1
Access: (0664/-rw-rw-r--)  Uid: ( 4157/   cshen)   Gid: ( 4157/   cshen)
Access: :06:
Modify: :08:
Change: :08:
Birth: -
[cisc361:/usa/cshen/ ] ls -i foo
73985 foo
[cisc361:/usa/cshen/ ] 
All info of each file is stored in the inode (persistent) structure

7. Soft/Symbolic Links
$ ln –s Chapter3 Chapter3.soft
Soft link is a file itself containing “pathname” for the file that the link file is a symbolic link to
3 files types
regular file (-)
directory (d)
symbolic link (l)

8. Symbolic (Soft) Links
A symbolic link is actually a file itself, of a different type, containing the pathname of the linked-to file
d: directory
-: regular file
l: symbolic link
Possible of dangling reference

9. Fundamental Issues
A, B, and C are located at the angles of an isosceles triangle
A lights up a torch
upon seeing the flare, B and C press buzzers
what would A, B, and C hear?
observers’ views of the global state of the system depend on the observation points
reason: non-instantaneous communications
One fundamental issue of distributed system is lack of a global system state
Two different watches
real clocks drift A B C
12:01PM
11:57AM
9:00AM
after 3 hours

10. Fundamental Issues
Since we cannot count on simultaneous observations of global states in distributed systems, we need to find a property on which we can depend
Distributed systems are causal
the cause precedes the effect
sending of a message precedes the receipt of the message

11. time
Space-Time diagram
p1 and r4?
p3 and q3?
concurrent
causal

12. time

13. Lamport Timestamps Example
Events occurring at three processors
local logical clocks are initialized to 0 2 3 7 1 6 4 3 1 5

14. Round Robin
Instead of running jobs to completion, RR runs a job for a (small) time slice (scheduling quantum) and switches to the next job in the ready queue; repeatedly does so until jobs are finished
Time slicing – length of time slice = multiple of timer-interrupt period
SJF response time = 5
RR response time = 1
Length of time slice is critical
under response time, the shorter the better
too short, overhead of context switching dominates

15. Incorporating I/O Process is blocked waiting for I/O completion
When I/O completes, an interrupt is raised, and OS runs and moves the process that issued the I/O from blocked state back to ready state
Poor use of CPU
Overlap CPU & I/O
higher CPU utilization
Interactive jobs get run frequently; while they are performing I/O, other CPU-intensive jobs run
job B has no I/O

16. Summary – History is the Guide
Rule 1: If Priority(A) > Priority(B), A runs (B doesn’t)
Rule 2: If Priority(A) = Priority(B), A & B run in RR
Rule 3: When a job enters the system, it is placed at the highest priority (the topmost queue)
Rule 4: Once a job uses up its time allotment at a given level (regardless of how many times it has given up the CPU), its priority is reduced (i.e., it moves down one queue)
Rule 5: After some time period S, move all the jobs in the system to the topmost queue
Instead of demanding a priori knowledge of a job, it instead observes the execution of a job and prioritizes it accordingly
It manages to achieve the best of both worlds: it can deliver excellent overall performance (similar to SJF/STCF) for interactive jobs, and is fair and makes progress for long-running CPU-intensive workloads

17. Working Set
When memory is simply oversubscribed (or memory demands of the set of running processes exceeds the available physical memory), the system will constantly be paging - thrashing
Working set
W(t, Δ) = the set of pages that a process has been referencing in the last Δ virtual time units
Admission control - given a set of processes, system decides not to run a subset of processes, with the hope that the reduced set of processes working sets (the pages that they are using actively) fit in memory and thus can make progress
time