Cherry cultivation has a long tradition in Bosnia and Herzegovina mainly due to favorable climatic conditions for cherry growing in this region. However, current cherry production is insufficient because of prevailing old cultivars and rootstocks. Modern intensive production of sweet cherry (Prunus avium L.) requires planting of high quality cultivars on dwarfing rootstocks planted in high density orchards. Cherry rootstock breeding programs worldwide require data on tolerance and performance of their rootstocks in different climatic conditions. Therefore, the influence of two cherry rootstocks ('Gisela 5 and 'Santa Lucia 64') on phenological events (blooming), growth and pomological properties of two cherry cultivars ('Stella' and 'Burlat') planted in modern orchard (managed according to standard commercial practice for integrated fruit production), near Sarajevo was evaluated. The trees grafted on 'Gisela 5' rootstocks were planted in 2004 and on 'Santa Lucia 64' in 2005. All tress were trained in a spindle system and the analyzed parameters were monitored in 2010. Rootstocks greatly influenced blooming time, growth and fruit characteristics of both investigated cultivars. The weaker rootstock was 'Gisela 5', which stimulated earlier blooming and caused statistically significant better fruit characteristics (fruit size, % flesh of fruit as well as total soluble solids content in fruit). The results of the analysis showed that both cherry cultivars reached better fruit quality on 'Gisela 5'. 'Stella' had better fruit quality than 'Burlat'. 'Santa Lucia 64' proved a better rootstock for 'Burlat' than for 'Stella'.

In this paper we present parallel implementation of genetic algorithm using map/reduce programming paradigm. Hadoop implementation of map/reduce library is used for this purpose. We compare our implementation with implementation presented in [1]. These two implementations are compared in solving One Max (Bit counting) problem. The comparison criteria between implementations are fitness convergence, quality of final solution, algorithm scalability, and cloud resource utilization. Our model for parallelization of genetic algorithm shows better performances and fitness convergence than model presented in [1], but our model has lower quality of solution because of species problem.