Final Flashcards
What does STS stand for
Short term scheduler
What does STS users care about
- Response Time
- Turnaround time
- Predictability
- Deadlines
What does STS systems care about
- Throughput
- CPU Utilization
- Fairness
- Priority
- Balancing Resources
What are two ways that STS processes decisions
– pre-emptive: a running process may be interrupted and moved from running to ready (timeslice interrupt)
– non pre-emptive: once a process is running, it continues until done, or blocked (for I/O
or OS service)
What polices can STS operate under
- First come, first served (FIFO)
- Round Robin
- Virtual Round Robin
- Shortest job first
- Shortest remaining time
What are the three types of message passing
- Blocking send, blocking recieve (good for tight sync)
- Non-Blocking send, blocking receive (generally most useful), good for client/server interactions
- Non-blocking send, non-blocking receive
What are two ways we identify sender and recevier in messaging?
Direct Addressing: Targeted, must know them.
Indirect addressing: Mailboxes, use common data structure
New class of scheduling for urgent processing
Real time scheduling. Usually these urgent processes are responding to events in the real world
Name the 7 layers in OSI
7 Application 6 Presentation 5 Session 4 Transport 3 Network 2 Data Link 1 Physical
What communicates on each layer of the OSI model?
Application: web applications Presentation: http Session: 80 Transport: TCP Network: IP/packets Data Link: Ethernet/frames Physical: Cat5
Real Time Scheduling must indicate timeliness, must have at least one of these
Deadlines:
Start time
End time
(rare it’s both)
Two types of real time scheduling deadlines
Hard real-time (must)
Soft real-time (should)
Real time system scheduling must by
preemptive and priority based
Real time scheduling must work in the following model for scheduling
Real time operating systems
What is deadlock
2 processes each waiting for the other to release a resource
-May be hardware or software
Why does deadlock occur?
Problem occurs when processes follow a request → use → release model of resource utilization
What is deadlock versus starvation
-Deadlock refers to the situation when
processes are stuck in circular waiting for the
resources.
-Starvation occurs when a process waits for
a resource indefinitely.
-Deadlock implies starvation but starvation
does not imply deadlock.
What are the conditions for deadlock
- Mutual exclusion (one process may use resource at a time)
- Hold and wait: (process hold allocated resources while waiting)
- No Pre-emption (no resource may be forcible taken)
- Circular wait (a closed chain of process where one holds what the other needs)
How can we prevent deadlock?
Prevention
Detection
Avoidance
What do we currently do to prevent deadlock
Nada… Let the OS figure it out or just ignore it as it rarely happens
Sequence of early programming
Op Codes (numbers only) Assembly Language (names for program, like add, mov, cmp)
What were the early program for translation
Translator: called assembler
Language called: Assembly Language
High level languages
Java, C, C++, Python…
What is software engineering
discipline devoted to best practices in design and
building of programs
Programming practies
- Write code easy to read
- Write code clear and easy to undertsand
- Comment code
- Write human readable form
What is the compiler
Translates a higher level language (human readable source code – Fortran, C, Cobol, etc.) to machine (object) code – the whole program at once before it is
loaded into memory.
What is the interpreter
Translates a higher level language (human readable source code – Fortran, C, Cobol, etc.) to machine (object) code – statement by statement at runtime
What is the assembler
Translates assembly language into machine (object) code.
What is the loader
loads load modules into memory:
What is the load module
Program that exist as a disk file, contains executable code
What is the linker
resolve such undefined references by linking library object code with our newly created object code
What are the different kinds of load modules
Absolute: program must be loaded into real memory at exactly the location it was built for
Relocatable: delay that association (binding) until load time
Dynamic Run-time:absolute addresses are calculated as needed at run time
What are two types of linking
Static and Dynamic
How do you get code into memory:
Basically, source code (1) gets translated into machine code (2) that goes to the linker (3). The linker than connects the machine code with preexisting libraries (4) creating the load module (5). The load module than “loads” the files into main memory (6).
How much does a sector hold
typically 512 bytes
How many sectors per track
63 on a hard disk
What is the difference between track and sector
Track is the circular portion, where the sector is the stops in a track
Cylinder-Head-Sector calc
cylinderheadsector * bytes per sector
i.e. 6296 * 16 * 63 * 512 bytes per sector
Average seek time per disk
3600: 8.5ms, Max is 20ms
7200: 4.17ms, Max 8.3
15000: 2ms, Max 4ms
Explain difference between CAV vs CLV
Bits in tracks at outer edge farther apart than those on inner tracks.
CAV: Constant angular velocity
CLV: Constant linear velocity
Explain delays fro optical disks
Worse, disk spins more slowly, seeds are slower, focus time on head
DVD layer change time
What makes SSD different
No mechenical parts, so these delays gone, no tracks
SSD are complicated
Writing is complicated, cells cannot be overwritten, erased then written.
SSD also have garbage collection (clean up fragmentation space )
What are two main disk control standards
SCSI: Small Computer Systems Interface
SATA: Serial ATA, Serial Advanced Technology Attachment
What is the difference in SCSI and SATA
- SCSI traditionally outperforms ATA, processor for controller
- SATA potentially surpasses SCSI on data transfer rates, possibly on overall performance, but uses more processor
