Week 3 Flashcards

1
Q

IN GFS: does master keep a persistant copy of the location of chunk servers

A

no

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

IN GFS: how does the master know where the chunk servers are?

A

the list is maintained via heartbeats so that the list will always be in sync with reality despite failures

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

In hadoop, how do you calculate the distance between 2 nodes

A

the sum of their distances to their closest common ancestor

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

what is the distance from proc1 to rack2

A
  1. It is how many hops between
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Do clients cache file data

A

no

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

does the chunk server cache file data

A

no, chunks are stored as local files, linux already keeps frequently accessed data in memory

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

what is strong consistancy

A

avoiding all inconsistances: no matter what identical

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What is update consistancy

A

When two people update same data at same time, without conflict handling, server will serialize them: apply one after the other

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

what is an ACID transaction, and is it valid for big data.

A

single transaction is:

Atomic

Cosistant

Isolated

Durable

NO!

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

CAP Theoream

A

Given Consistancy, Avalilablity, and Partition Tolerance, you can only get 2

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

what is consistancy

A

every read recieved the most recent write, or an error

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

what is availablity

A

every request recieved a non error response, without guarentee that it contains the most recent write

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

what is partition tolerance

A

the system continues to operate despite an arbitrarty number of messages being dropped or delayed by network between nodes

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

what does eventuall consistant mean>?

A

at any time nodes may have replication inconsistances.

if there are no more updates, eventually all nodes will be updated to the same value

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What does a mutation do

A

changes the content or metadata of a chunk

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

IN GFS: what does the state of a file region after mutation depend on

A

Type of mutation

Sucess or failure of mutation

whether there was concurrent mutations

17
Q

GFS has relaxed consistanty model, what does this mean?

A

See the same data on all replicas from users point of view

18
Q

what is this, each of these are diff chunk servers

A

inconsistant and defined

19
Q

what is this, each of these are diff chunk servers

A

consistant but undefined

20
Q

what is this, each of these are diff chunk servers

A

defined

21
Q

what is the file state after a write and a serial success

A

defined

22
Q

what is the file state after a record append and a serial sucess

A

defined or possibly inconsistant

23
Q

what is the file state after a concurrent sucess and record append

A

defined or possibly inconsistant

24
Q

what is the file state after a concurrent sucess and write

A

consistant but undefined

25
Q

what is the filestate after a failed write or append

A

Inconsistant

26
Q

TF: GFS guarantees the mutated file to be defined and to contain the data written by the last mutation

A

T

27
Q
A