Skip to main content
Fig. 6 | BMC Bioinformatics

Fig. 6

From: K-mer clustering algorithm using a MapReduce framework: application to the parallelization of the Inchworm module of Trinity

Fig. 6

Data objects created by MapReduce-Inchworm for the experimental mouse dataset (see Table 1). (a) The size of data objects generated by each MapReduce function on each compute node, as calculated from the number of KV/KMV pairs involved. The data sizes for each MapReduce function (map(), collate(), and reduce()) were averaged over nodes and iterations within each of the 5 main steps of MapReduce-Inchworm. (b-e) The corresponding cumulative I/O to disk, due to out-of-core processing, per compute node. Results are shown with pagesize set to: (b) 1 GB, (c) 2 GB, (d) 3 GB and (e) 4 GB. For all graphs, the Y-axis gives the data size in GB

Back to article page