CS848 Paper Review Form - Fall 2006

Paper Title: Continuous resource monitoring for self-predicting DBMS

Author(s): D. Narayann, E. Thereska and A. Ailamaki

1) Is the paper technically correct?
 [X] Yes
 [ ] Mostly (minor flaws, but mostly solid)
 [ ] No

2) Originality
 [ ] Very good (very novel, trailblazing work)
 [X] Good 
 [ ] Marginal (very incremental)
 [ ] Poor (little or nothing that is new)

3) Technical Depth
 [ ] Very good (comparable to best conference papers)
 [X] Good (comparable to typical conference papers)
 [ ] Marginal depth
 [ ] Little or no depth

4) Impact/Significance
 [ ] Very significant
 [X] Significant
 [ ] Marginal significance.
 [ ] Little or no significance.

5) Presentation
 [ ] Very well written
 [X] Generally well written
 [ ] Readable
 [ ] Needs considerable work
 [ ] Unacceptably bad

6) Overall Rating
 [ ] Strong accept (very high quality)
 [X] Accept (high quality - would argue for acceptance)
 [ ] Weak Accept (marginal, willing to accept but wouldn't argue for it)
 [ ] Weak Reject (marginal, probably reject)
 [ ] Reject (would argue for rejection)

7) Summary of the paper's main contribution and rationale
   for your recommendation. (1-2 paragraphs)

This paper presents a tool for monitoring database resource usage and providing
a tool for predicting the behaviour of the DBMS under changes in load.  This
problem is important because effective resource provisioning is necessary to avoid
costly over-provisioning of resources.  Typically, the number of resources assigned
to an application exceeds the required amount.  This tool looks at specific
charateristics in the system to predict the amount of resources needed in the future.

They system uses several criteria measured by tracing execution of queries and
transactions in the database.  These criteria include control flow, cpu scheduling,
buffer pool activity and disk io.  They look at two performance metrics, the 
throughput and the response time.  They generate several formulas to formalize
these metrics and verify that the formulas can predict performance through experiments.



8) List 1-3 strengths of the paper.  (1-2 sentences each,
identified as S1, S2, S3.)

S1 - Looks at a wide variety of characteristics when making predictions.

S2 - Provides an interface easily understood by humans.

S3 - Allows predictions to be made on the fly.

9) List 1-3 weaknesses of the paper (1-2 sentences each,
identified as W1, W2, W3.)

W1 - Experiments and implementation lack completeness.

W2 - The end of the paper seems rushed and it feels like certain aspects were
left out because of time constraints.  It would be a better paper with these
measurements.

W3 - It was not made explicitly clear why only considering OLTP type workloads
was well justified.

10) Detailed comments for authors.

Overall, it was a good paper that had an unfinished feel.  Some more experiments using
a wider variety of different resources would have been appreciated.