US 2014/0280386 A1 · Persisting Large Volumes of Data in an Efficient, Unobtrusive Manner

Abstract

Embodiments provide a data persisting mechanism that allows for efficient, unobtrusive persisting of large volumes of data while optimizing the use of system resources by the persisting process. In an embodiment, the persisting process includes a self-tuning algorithm that constantly monitors persistence performance and that adjusts persistence time to maintain performance within user-defined criteria. From one aspect, this allows the persisting process to seamlessly adapt to changes in system environment (speeding up persistence during times of low processor usage and slowing down persistence during times of high processor usage) and to reduce or eliminate CPU spikes caused by persisting process. From another aspect, the persisting process results in the data being persisted as quickly as possible given the system constraints, thereby minimizing the possibility of data loss.

Claims20

Dependent claims

A system, comprising:

a storage;
and
a processor, coupled to the storage, configured to execute an application, wherein the application is configured to:
determine, based on a current performance level of the system, a first persistence data rate for persisting a plurality of objects to the storage;
if a persistence time required to persist the plurality of objects to the storage at the first persistence data rate is within a persistence cycle time, persist the plurality of objects to the storage at the first persistence data rate;
and
if the persistence time required to persist the plurality of objects to the storage at the first persistence data rate exceeds the persistence cycle time,
increase the first persistence data rate to generate a second persistence data rate;
and
persist the plurality of objects to the storage at the second persistence data rate.

7 more elements

A computer program product comprising a non-transitory computer readable medium having control logic stored therein, the control logic, when executed by a processor, causes the processor to perform a method for persisting data to a storage responsive to system performance, the method comprising:

determining, based on a current system performance level, a first persistence data rate for persisting a plurality of objects to the storage;
if a persistence time required to persist the plurality of objects to the storage at the first persistence data rate is within a persistence cycle time, persisting the plurality of objects to the storage at the first persistence data rate;
and
if the persistence time required to persist the plurality of objects to the storage at the first persistence data rate exceeds the persistence cycle time,
increasing the first persistence data rate to generate a second persistence data rate;
and
persisting the plurality of objects to the storage at the second persistence data rate.

4 more elements

Dependencies are inferred from the claim text.

Persisting Large Volumes of Data in an Efficient, Unobtrusive Manner

Life of this patent

Persisting Large Volumes of Data in an Efficient, Unobtrusive Manner

Life of this patent

Abstract

Claims20