Resource disaggregation-based DCs may face severe performance degradations for VM-to-VM communications due to their impact on network utilization. We analyse this phenomenon and investigate how it relates to blade and/or network resource dimensioning for a proper compensation of the effect