Optimizing big data processing performance in the public cloud: opportunities and approaches
Bibliographic record
Abstract
Today's lightning fast data generation from massive sources is calling for efficient big data processing, which imposes unprecedented demands on the computing and networking infrastructures. State-of-the-art tools, most notably MapReduce, are generally performed on dedicated server clusters to explore data parallelism. For grass roots users or non-computing professionals, the cost of deploying and maintaining a large-scale dedicated server clusters can be prohibitively high, not to mention the technical skills involved. On the other hand, public clouds allow general users to rent virtual machines and run their applications in a pay-as-you-go manner with ultra-high scalability with minimal upfront costs. This new computing paradigm has gained tremendous success in recent years, becoming a highly attractive alternative to dedicated server clusters. This article discusses the critical challenges and opportunities when big data meet the public cloud. We identify the key differences between running big data processing in a public cloud and in dedicated server clusters. We then present two important problems for efficient big data processing in the public cloud, resource provisioning (i.e., how to rent VMs) and VM-MapReduce job/task scheduling (i.e., how to run MapReduce after the VMs are constructed). Each of these two questions have a set of problems to solve. We present solution approaches for certain problems, and offer optimized design guidelines for others. Finally, we discuss our implementation experiences.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.005 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.003 |
| Science and technology studies | 0.002 | 0.002 |
| Scholarly communication | 0.007 | 0.010 |
| Open science | 0.003 | 0.003 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".