Click here to Skip to main content
15,889,281 members
Home / Discussions / Algorithms
   

Algorithms

 
GeneralRe: text diff Pin
m-khansari24-Dec-08 22:55
m-khansari24-Dec-08 22:55 
QuestionRadian angle to Quaternion Pin
shaibee18-Dec-08 3:29
shaibee18-Dec-08 3:29 
AnswerRe: Radian angle to Quaternion Pin
73Zeppelin18-Dec-08 3:42
73Zeppelin18-Dec-08 3:42 
GeneralRe: Radian angle to Quaternion Pin
shaibee18-Dec-08 4:01
shaibee18-Dec-08 4:01 
GeneralRe: Radian angle to Quaternion Pin
73Zeppelin18-Dec-08 4:48
73Zeppelin18-Dec-08 4:48 
QuestionPseudocode implementation of the strict timestamp protocol for ensuring serializability, recoverability and cascadeless roll-backs [modified] Pin
efgtdg17-Dec-08 10:07
efgtdg17-Dec-08 10:07 
AnswerRe: Pseudocode implementation of the strict timestamp protocol for ensuring serializability, recoverability and cascadeless roll-backs Pin
73Zeppelin17-Dec-08 22:17
73Zeppelin17-Dec-08 22:17 
GeneralRe: Pseudocode implementation of the strict timestamp protocol for ensuring serializability, recoverability and cascadeless roll-backs Pin
efgtdg18-Dec-08 0:39
efgtdg18-Dec-08 0:39 
I apologize for not providing the question.

The question is:

Is this pseudocode correct regarding the strict time stamp ordering protocol ?
( Strict TS protocol is used for ensuring conflict serializability, recoverability and cascade less roll-backs )

Some background knowledge ( I hope this to clarify things):

1 ) Conflict serializability -> This is ensured by the basic time stamp protocol
2 ) But the basic TS protocol does not ensure (1) recoverability and (2) cascadeless roll-backs.
3 ) This is were "strict" steps in and ensures the aforementioned 2 features.

e.g. Transactions TS(T1)<TS(T2)<TS(T3)

Assume that TS(T2) has performed a write_item(X) and has committed.
Additionally, let us assume that TS(T3) performed the following 2 operations after the previous commit:
Y=Y+X
write_item(Y)
but has not committed yet.

Then, transaction T1 does attempt to perform a write_item(X).

The basic time stamp protocol will force transaction T1 to abort (roll-back), because TS(T1) < write_TS(X), write_TS(X) == TS(T2)

However, since the transaction T2 has committed it is not recoverable through a roll-back anymore.
Furthermore, transaction T3 has to be rolled-back as well. If another transaction T4 had based its writes on T3 then it should also be rolled-back. This is were cascading roll-backs happen.

Thanks, for using your time to answer my question.
GeneralRe: Pseudocode implementation of the strict timestamp protocol for ensuring serializability, recoverability and cascadeless roll-backs Pin
73Zeppelin18-Dec-08 2:00
73Zeppelin18-Dec-08 2:00 
GeneralRe: Pseudocode implementation of the strict timestamp protocol for ensuring serializability, recoverability and cascadeless roll-backs Pin
efgtdg19-Dec-08 0:51
efgtdg19-Dec-08 0:51 
GeneralRe: Pseudocode implementation of the strict timestamp protocol for ensuring serializability, recoverability and cascadeless roll-backs Pin
73Zeppelin19-Dec-08 2:30
73Zeppelin19-Dec-08 2:30 
QuestionQuestion from my Phone Interview Pin
BobInNJ15-Dec-08 10:17
BobInNJ15-Dec-08 10:17 
GeneralSum Pin
Luc Pattyn15-Dec-08 11:19
sitebuilderLuc Pattyn15-Dec-08 11:19 
AnswerRe: Question from my Phone Interview Pin
Member 419459315-Dec-08 11:47
Member 419459315-Dec-08 11:47 
AnswerRe: Question from my Phone Interview Pin
Member 419459315-Dec-08 11:49
Member 419459315-Dec-08 11:49 
GeneralRe: Question from my Phone Interview Pin
BobInNJ15-Dec-08 12:43
BobInNJ15-Dec-08 12:43 
GeneralRe: Question from my Phone Interview Pin
Luc Pattyn15-Dec-08 12:52
sitebuilderLuc Pattyn15-Dec-08 12:52 
GeneralRe: Question from my Phone Interview Pin
Member 419459315-Dec-08 13:05
Member 419459315-Dec-08 13:05 
Generalsum &lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt; Pin
Luc Pattyn15-Dec-08 13:20
sitebuilderLuc Pattyn15-Dec-08 13:20 
GeneralRe: sum &lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt; Pin
Member 419459315-Dec-08 14:43
Member 419459315-Dec-08 14:43 
GeneralRe: sum &lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt;&lt; Pin
riced18-Dec-08 12:47
riced18-Dec-08 12:47 
GeneralNot Cobol Pin
Luc Pattyn18-Dec-08 12:59
sitebuilderLuc Pattyn18-Dec-08 12:59 
GeneralRe: Not Cobol Pin
riced18-Dec-08 13:27
riced18-Dec-08 13:27 
GeneralRe: Not Cobol Pin
Luc Pattyn18-Dec-08 13:39
sitebuilderLuc Pattyn18-Dec-08 13:39 
GeneralRe: Question from my Phone Interview Pin
leonej_dt15-Dec-08 13:33
leonej_dt15-Dec-08 13:33 

General General    News News    Suggestion Suggestion    Question Question    Bug Bug    Answer Answer    Joke Joke    Praise Praise    Rant Rant    Admin Admin   

Use Ctrl+Left/Right to switch messages, Ctrl+Up/Down to switch threads, Ctrl+Shift+Left/Right to switch pages.