LARGE SCALE STORAGE SYSTEM AND METHOD OF OPERATING THEREOF

摘要

A distributed storage system comprising interconnected computer nodes; each one of the computer nodes comprising at least one processing resource configured to execute a Unified Distributed Storage Platform (UDSP) agent; at least one of the computer nodes comprising one or more resources including at least one cache resource configured to cache objects and having corresponding cache-related parameters; at least one UDSP agent of a respective computer node having the at least one cache resource is configured to: monitor cache-related parameters of the at least one cache resource connected to the respective computer node, for determining whether the cache-related parameters meet at least one first SLS criteria; and in the case the at least one first SLS criteria is not met, initiate handoff of at least part of one or more cache object spaces of the at least one cache resource to at least one other computer node, which after receiving the at least part of one or more cache object spaces, its cache-related parameters meet at least one second SLS criteria.

主权项

1. A distributed storage system comprising interconnected computer nodes;
each one of said interconnected computer nodes comprising at least one processing resource configured to execute a Unified Distributed Storage Platform (UDSP) agent; wherein a first computer node among said interconnected computer nodes comprises a first cache resource, said first computer node being responsible for handling one or more cache object spaces related to one or more logical storage entities; and wherein a UDSP agent of the first computer node is configured to:
monitor cache-related parameters corresponding to said first cache resource to determine whether the cache-related parameters meet at least one first Service Level Specification (SLS) criterion;when said at least one first SLS criterion is not met, initiate a handoff of responsibility over at least part of said one or more cache object spaces to at least one second computer node among said interconnected computer nodes, said second computer node comprising a second cache resource, wherein said UDSP agent is further configured to select said at least one second computer node such that, after receiving responsibility over said at least part of said one or more cache object spaces, cache-related parameters corresponding to said second cache resource meet at least one second SLS criterion;create a post-handoff cache mapping at least indicative of said at least one second computer node selected to be responsible for said at least part of said one or more cache object spaces;send said post-handoff cache mapping to said at least one second computer node; andtransfer responsibility over said at least part of said one or more cache object spaces to said at least one second computer node.