دانلود Self managing monitoring for highly elastic large scale cloud deployments

Abstract

Infrastructure as a Service computing exhibits a number of properties,  which  are  not  found  in  conventional  server  deployments.  Elasticity is among the most significant of these properties  which  has  wide  reaching  implications  for  applications deployed in cloud hosted VMs.  Among the applications affected by elasticity is monitoring. In  this  paper  we  investigate  the  challenges  of  monitoring large cloud deployments and how these challenges differ from previous monitoring problems.  In order to meet these unique  challenges  we  propose  Varanus1,  a  highly  scalable monitoring  tool  resistant  to  the  effects  of  rapid  elasticity. This tool breaks with many of the conventions of previous monitoring systems and leverages a multi-tier P2P architecture in order to achieve in situ monitoring without the need for dedicated monitoring infrastructure. We then evaluate Varanus against current monitoring architectures.  We find that conventional monitoring tools per-form acceptably for small, non changing cloud deployments. However in the case of large or highly elastic deployments current tools perform unacceptably incurring increased latencies,  high  load  and  slowed  operation  necessitating  that a  new,  alternative  tool  be  used.   Further,  we  demonstrate that Varanus maintains low latency monitoring with limited demand upon resources, even during during periods of high elasticity.