XML CO527 Anonymous Questions and Answers

This page lists the various questions and answers. To submit a question, use the anonymous questions page. You may find the keyword index and/or top-level index useful for locating past questions and answers.

We have taken the liberty of making some minor typographical corrections to some of the questions as originally put. Although most of the questions here will have been submitted anonymously, this page also serves to answer some questions of general interest to those on the course.

Question 21:

Submission reference: IN3629

Are you looking for the GMT or the GMT+1 timezone in regards to question 3 of Assessment 4?

Answer 21:

I'm looking for what it actually says in the directory.

Question 22:

Submission reference: IN3676

"And to convert a cluster number (which is what appears in the root directory) to a block number, we need to add 0x17, to allow for that strange offset of 2." From the fat tutorial, I don't understand where this offset of 2 comes from? We've taken 2 away from 0x19 which is where the first allocation unit starts, but why?

Answer 22:

Clusters are numbered starting at 2. See: http://www.tavi.co.uk/phobos/fat.html#file_allocation_table, in particular the text in red.

Reason? The first two entries in the FAT were historically used for something else.

Question 23:

Submission reference: IN3678

Is the method for question 6 similar to question 5?

Answer 23:

Yes, but you have to do more work.

Question 24:

Submission reference: IN3682

On what date were most of the files created? [4] For this, do i simply analyse all the dates of the files and then count the majority or is there some other (More Efficient) way?

Answer 24:

Yes, analyse all the dates and just count them. It's not a big job.

Question 25:

Submission reference: IN3683

Regarding question 2 of Assessment 4, does a disk volume label count as a root directory entry, or do the directory entries count as the files themselves? Eg. CALENDAR.TXT as the first directory entry.

Answer 25:

The volume label occupies a slot in the root directory. Thus it counts as an entry.

Question 26:

Submission reference: IN3681

If we answer more than the specified questions in the exam, how will the marks be worked out please?

Answer 26:

The general policy is that we'll mark all questions answered and give you the best mark possible under the rubric. However, the time limit in the exam means that you should probably aim to answer fewer questions well, rather than all questions mediocre/poorly. Section A of the exam contains 50 marks, but will be capped at 40 marks regardless. Section B contains 3 questions of which you should answer 2. These are 20 mark questions, so better to do two of these well, rather than 3 badly (since you'll only get the marks for the best 2).

Keywords: exam

Question 27:

Submission reference: IN3684

How concise should our explanations for filesystems be? For example, my answer to q5 is about 11 lines long.

Answer 27:

The answer needs to show all the working you used to get the answer. You don't have to give details of every binary to decimal conversion, etc., but you do need to show you haven't just plucked the values out of the air.

I would think that 5 lines might be a bit short, but you may prove me wrong!

Question 28:

Submission reference: IN3689

Do you know of any good resources for learning about C-SCAN? I have tried googling it, but the results are all just uncommented code which aren't particularly helpful at all.

Answer 28:

The best reference (give or take) is likely to be an OS textbook: check the library!

Question 29:

Submission reference: IN3690

I feel like I dont quite understand question 7 of the filesystems assessment. It seems the exact same as q6 but you have to count to 7 instead of 3? yet its worth 4 more marks?

Answer 29:

It is actually quite a bit harder...

Question 30:

Submission reference: IN3691

I feel like I dont quite understand question 7 of the filesystems assessment. It seems the exact same as q6 but you have to count to 7 instead of 3? yet its worth 4 more marks?

Answer 30:

Answered before...but question 7 is quite a bit harder.

Question 31:

Submission reference: IN3692

i've prefaced all values and offsets in the filesystems assessment with 0x, I assume thats either correct or wont get marked down? If I recall correctly it denotes an integer?

Answer 31:

The prefix 0x means that the number is expressed in hexadecimal. This is a common notation in a number of programming languages.

Question 32:

Submission reference: IN3693

I really don't understand why q7 is harder. You use the exact same method for both questions to get to the FAT and then you just find the 7th entry?

Answer 32:

Let's say then that some people find it harder. I can't say more than that without giving things away.

Question 33:

Submission reference: IN3707

For the final assessment, are we allowed to use functions from Java.Math?

Answer 33:

No.

Keywords: disk-scheduling

Question 34:

Submission reference: IN3712

Can you provide a detailed description on how to run the disk scheduler as I don't know how to run it from your instructions. Thanks.

Answer 34:

Not really.. Probably best to ask one of your peers to show you, or pop along to my office (email to arrange a time perhaps) and I can go through it with you quickly. The instructions provided should be sufficient however — what was the specific problem you encountered?

Question 35:

Submission reference: IN3709

How many marks is it possible to gain by implementing SSTF?

Answer 35:

This depends largely on the submissions overall, but somewhere in the range 55% to 65%.

Keywords: disk-scheduling

Question 36:

Submission reference: IN3714

I have no idea where to start with disk scheduler. I know I want to implement C-scan but that's about it.

Answer 36:

First step, understand the code that is there already (i.e. the FIFO implementation). The slides I made should help somewhat here, but you may need to refer to textbooks and the like as well (depending on what you do or don't understand).

Keywords: disk-scheduling

Question 37:

Submission reference: IN3716

Can I implement the C-Look algorithm?

Answer 37:

You can implement whatever algorithm you choose! (the proviso being that it works and hopefully performs better than FIFO and/or SSTF).

Keywords: disk-scheduling

Question 38:

Submission reference: IN3731

Are we allowed to use "Math.abs"?

Answer 38:

No. But you are (obviously) allowed to use absolute value functionality, you just have to implement it yourself.

Keywords: disk-scheduling

Question 39:

Submission reference: IN3717

I'm trying to implement SSTF. I think I've pretty much done it, but I don't really understand how to keep the most recently read block. At least, I'm guessing thats whats going wrong.

When read_complete(blk, BRequest) is called, is blk the block the head is currently at? As far as I can tell I've implemented each step in the slides about the problem otherwise.

Answer 39:

If "blk" is >= 0 then yes, it's the block the head is currently on (i.e. the requested block has been read, hence read complete). If "blk" is < 0 then it's as stated in the comments in the code (i.e. idle and able to accept a new request). This means you can't rely on "blk" being "the most recently read block" universally, only when it is >= 0.

Keywords: disk-scheduling

Question 40:

Submission reference: IN3718

I keep on getting unexpected blocks while trying to implement SSTF. Is there any typical problem I should look out for?

Answer 40:

This suggests something is wrong somewhere — I'd recommend creating some simple test-case data and/or putting in a bunch of debugging and/or running with "-v" (verbose). The specific error ("unexpected block") is because you're telling the high-level code ("highlevel_didread") about a block it didn't ask for. In practical terms, because blk or req passed to it are defunct; there may be various reasons for this, but that's where debugging and whatnot can help.

Keywords: disk-scheduling