An Analysis of Traces from a Production MapReduce Cluster
By Soila Kavulya, Jiaqi Tan, Rajeev Gandhi, Priya Narasimhan. 2010 10th IEEE/ACM International Conference on Cluster, Cloud and Grid Computing, 2010
362 citations recorded by the source corpus.
Subfields: Computer Networks and Communications, Information Systems
View Paper: https://www.semanticscholar.org/paper/f8e8e355ae3e6bcca65a8da99fee831565cce0cf
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computer Networks and Communications
- Shivaram Venkataraman
- Michael Abd-El-Malek
- Roy H. Campbell
- John Wilkes
- L. Cherkasova
- Malte Schwarzkopf
- Andy Konwinski
- Myeongjae Jeon
In Information Systems
- Shivaram Venkataraman
- Michael Abd-El-Malek
- Roy H. Campbell
- John Wilkes
- L. Cherkasova
- Malte Schwarzkopf
- Andy Konwinski
- Fan Yang
Citations in the corpus
- Multi-tenant GPU Clusters for Deep Learning Workloads: Analysis and Implications (2018, background): “conducted a detailed characterization for job failures in a production MapReduce cluster [30].”
- ARIA: automatic resource inference and allocation for mapreduce environments (2011, methodology): “(The study in [10] did not report the statistics for individual map and reduce task durations, so we chose them as described above).”
- Analysis of Large-Scale Multi-Tenant GPU Clusters for DNN Training Workloads (2019, background): “However as opposed to prior study on big data traces [30], we see a significant fraction of jobs (30.7% as shown in Table 6) are either terminated unsuccessfully or killed by users.”
- Omega: flexible, scalable schedulers for large compute clusters (2013, result): “These results are broadly similar to other analyses of cluster traces from Yahoo [17], Facebook [7] and Google [20, 24, 25, 29].”