当前位置:天才代写 > CS代写,CS作业代写考试代考-价格便宜有靠谱 > 深度学习作业代写 Raid DP代写

深度学习作业代写 Raid DP代写

2022-02-14 11:40 星期一 所属: CS代写,CS作业代写考试代考-价格便宜有靠谱 浏览:514

Assignment 2 Raid DP Rebuild

深度学习作业代写 Assignment 2 Raid DP Rebuild I will also be posting a recording showing the initial steps of implementing each “level” of the solution.

I will also be posting a recording showing the initial steps of implementing each “level” of the solution.
Finally, I will be posting a second version of the raid_dp.ipynb notebook containing the chnge I make during the recording.

Download the “raid_dp.ipynb” notebook used in this evening’s lecture.

It is linked in the Course Materials, as well in this assignment.

Install a copy of Jupyterlab, or identify a publicly available copy and upload the notebook to it.

Find the “rebuild_drives” function definition.

Note the “else” clause, rebuilding two data drives (or one data drive and the Raid-4 Parity) form the diagonal parity drive, only prints “I ran out of time!”

 

深度学习作业代写
深度学习作业代写

 

See if you can figure out how to rebuild two arbitrary data/parity drives from the diagonal parity, using the hints in the code comments.

Remember, for any given RAID-DP “diagonal slice”:
The segment number (within the extent) on the dp drive (drives[-1]) for slice N is segment N.
The drive at (N-1) modulo (number_of_data_drives + 1) is NOT a part of slice N.
For every drive in slice N, the drive_number modulo (number_of_data_drives+1) is N greater than the block(segment) that is part of the slice. (for slice 2, drive 3 block 1 is part of the slice, because 3-1=2)

These rules should allow you to choose the first two blocks to rebuild easily. After that, it’s a matter of chaining the horizontal and diagonal rebuilds in the correct order.

Consider creating a map remembering which blocks are currently available, and which are unavailable.

 

PARTIAL CREDIT NOTES  深度学习作业代写

If you simply add (and run) a hard-coded rebuild of a single extent on drives 0 and 1 (as described in comments in the code), that will be a level of partial credit.
If you successfully rebuild all extents on the two fixed drives, that is more partial credit.
If you successfully rebuild two arbitrary drives, but hardcode the number of drives and blocks, that is more partial credit.
If you successfuly rebuild two arbitrary drives given an arbitrary number of drives and data blocks, that is full credit.

Modified raid_dp_01.ipynb, with a *slightly* cleaner internal API (who am I kidding, it’s using global variables) has been posted. Please consider using this version for your starting point.

 

2021-01-31  深度学习作业代写

As promised, a new version of the notebook, raid_dp_02.ipynb, is uploaded.
Instead of failing drives 2 and DP, it fails drives 0 and 1
It partially “hard rebuilds” data in the first extent of drives 0 and 1.
From this, you should be able to easily hard rebuild all data on drives 0 and 1. This is worth partial credit, if submitted.
The code also contains comments giving ‘hints’ on some of the helper functions needed to implement a generic solution for any 2 drives.
Remember, there are unlimited submissions for this assignment. If you want to hard-code drives 0 and 1, and then resubmit the generic solution, feel free.

 

2021-02-03

Yet another version, raid_dp_03.ipnb, is uploaded, fixing a bug in the diagonal parity block rebuild

 

深度学习作业代写
深度学习作业代写

 

 

 

更多代写:高分CS Quiz代写  gre助考  管理学英国代写   法学毕业论文代写  管理类assignment代写  代写本科毕业论文

合作平台:essay代写 论文代写 写手招聘 英国留学生代写

 

天才代写-代写联系方式