3.7 KiB
-
-
Review **Technical Innovation ** * [ ] 1 - Routine work, untested technical work or impractical idea * [ ] 2 - Good work, not particularly novel, akin to a routine evolution of existing technologies * [x] 3 - Good technical work with some novel features described * [ ] 4 - Very innovative technical work that demonstrates clear thought leadership for HPE * [ ] 5 - Clearly a breakthrough with significant technical innovation
Business Impact * [ ] 1 - Impractical idea; limited business value * [ ] 2 - Good work, but with limited direct or indirect business value, no clear path to capture business value+ * [x] 3 - Moderate business impact that merits further assessment * [ ] 4 - Work will provide HPE with valuable and meaningful differentiation in the market * [ ] 5 - Clearly and significant impacts HPE’s business, opens new market opportunities
Clarity of Presentation * [ ] 1 - Difficult to understand; confusing; incomplete description; very short * [ ] 2 - Hard to follow; uses unfamiliar terminology or acronyms; missing important data * [ ] 3 - Understandable but lacking some relevant information * [x] 4 - Clear and logical; some important information is missing or unclear * [ ] 5 - Very clearly described; logical flow; well supported with practical results and proof points
Overall Recommendation * [ ] 1 - Reject * [ ] 2 - Weak Reject * [ ] 3 - Weak Accept * [x] 4 - Accept * [ ] 5 - Strong Accept
Suggested Presentation Style What type of presentation do you recommend for this submission? * [x] Formal Session * [ ] Poster Session * [ ] Not recommended for presentation
Favorite * [X] No * [ ] Yes
Reviewer Confidence * [ ] 1 - No confidence - I am not qualified to pass judgement on this submission * [ ] 2 - Low confidence - I do not have enough experience in this area to make a definitive decision on this submission * [ ] 3 - Somewhat confident - I have a reasonable understanding of this research area * [x] 4 - Confident - I have considerable confidence in my review and understanding of this work * [ ] 5 - Very Confident - I am confident in my review and understanding of the work
Comments for the Authors Provide constructive comments to the author(s).
-
The author(s) outline a challenge associated with FaaS/Container-based computing, where there is significant overprovisioning of resources both on a per process and cluster-wide scale. The solution proposed looks to provide a predictive scaling system (instead of the traditional load/reactive scalers).
-
I have to applaud the author(s) scientific and experimental approach to this problem - I found the experimental data in the evidence section fascinating, and it showed a healthly level if self-critique of the approach.
-
Overall I think the approach is sound and when coupled with sensible cool-down timers, lower limits, etc could provide significant benefits in large scale systems.
-
While the paper could be clearer, this is a good idea that deserves further exploration. I would be interested to see an exploration of how this services scales and any latency this may introduce. Comments for the Program Committee (authors will not see these comments) Provide comments to the PC (if any) that should not be shared with the author(s). An interesting paper. I think the author(s) did a great job of balancing the amount of detail given about the proposal whilst providing significant space to the evidence section. I would love to see more papers that show significant evidence (where available) instead of the single paragraph most provide - can we update the submission guidance to encourage this?
