CS848 Paper Review Form - Fall 2006

Paper Title: Performance Guarantees for Web Server End-Systems: A Control-Theoretical Approach

Author(s): T. Abdelzaher, K. Shin, N. Bhatti

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
 [ ] Generally well written
 [x] 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)

The paper deals with bandwidth allocation and QoS adaption in Web servers.
Control theory is applied to control performance of a Web server.
A utilization control loop is used to satisfy pre-defined bound and individual
time constraints. Extensions are demonstrated to provide performance isolation,
service differentiation, excess capacity sharing and QoS guarantees.
To evaluate the results, an implementation using the Apache Web server is presented
as well.

The paper gives a good explanation of the control theory being used. It clearly 
defines its goals and the way the actuators and sensors are to be designed and implemented.
The performance evaluation of the developed software discusses the necessary topics in 
detail and the performed analyzis brings more value to the paper.

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

S1. Achieves performance isolation, service differentiation, excess capacity sharing.
S2. Well-defined and implemented control theory.
S3. Extensive testing and analysis performed.


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

W1. Static web content only.
W2. Using a "magic" number (U < 0.58)
W3. Single pre-defined bottleneck.

10) Detailed comments for authors.

Since the bottleneck is the connection bandwidth, rather than just having admission control
(concurrent connections) and QoS degradation (i.e. lower quality images), what about 
throttling of the transfer rate of requested files, i.e. some sort of packet scheduling?