Light-weight operation of a failover system for Cloud computing

Abstract

Availability is one of the biggest challenges which slow down the adoption of Cloud computing in the IT industry. A failover system can be designed to use as a backup in case of Cloud failure or unavailability. In this paper, we introduce the method for designing a local failover system for Cloud. Since a failover system has less resource and limited scalability; it cannot handle all workloads previously on the Cloud. We overcome this drawback by adapting the full operation performed in the Cloud into a light-weight operation optimized for a failover system. A light-weight operation consumes less resource but also has reduced quality of service (QoS) compared to the full operation. We also define a write type of data according to change and future use of that data. We adapt the full operation into a lightweight one by reducing the number of write types or changing one type to another. In other words, we reduce the quality of service of the system in order to serve more requests. Experimental result shows that a system with full operation can sustain an average of 472 maximum numbers of requests, and a system with light-weight operation can sustain 844 requests, an improvement of 179%.

Thaiwrittennet: Thai Handwritten Script Recognition Using Deep Neural Networks

“…Afterward, ConvNet was proven to be more accurate than a handcraft feature with traditional machine learning. On the dark side, ConvNet was used to train the spamming bots to recognize those reCaptcha images and break the human verification (Mookdarsanit & Mookdarsanit, 2020a) against authentication (Soimart & Mookdarsanit, 2016b) mechanism that finally made the sever-side system processed a large number of junk jobs as the concurrent workloads (Mookdarsanit & Gertphol, 2013). ConvNet was designed to have highly optimized structures (Mookdarsanit & Mookdarsanit, 2019a) to learn the extraction and abstraction of 2D features.…”

Location Estimation of a Photo: A Geo-signature MapReduce Workflow

“…Customarily, the purpose of "MapReduce approach" [51][52][53] was designed for the actual "dynamic structure" for the "velocity", "volume" and "variety" of non-volatile large-scale datasets (or Big data [1]). In case of a geo-untagged photo, MapReduce indexing filters only the useful geo-tagged photos (that were collected in term of vectors with geo-tagging) from the set of large-scale samples which are similar to some visual contents of geo-untagged photo.…”