Adaptive multi-level checkpointing
AssigneeHewlett Packard Enterprise Development LP
InventorsCong Xu, Itir Akgun, Paolo Faraboschi
Life of this patent
Estimating the term…
Abstract
In some examples, with respect to adaptive multi-level checkpointing, a transfer parameter associated with transfer of checkpoint data from a node-local storage to a parallel file system may be ascertained for the checkpoint data stored in the node-local storage. The transfer parameter may be compared to a specified transfer parameter threshold. A determination may be made, based on the comparison of the transfer parameter to the specified transfer parameter threshold, as to whether to transfer the checkpoint data from the node-local storage to the parallel file system.
Claims18
An apparatus comprising:
- a processor;
- and
- a non-transitory computer readable medium storing machine readable instructions that when executed by the processor cause the processor to:
- ascertain, for checkpoint data stored in node-local storage, a transfer parameter associated with transfer of the checkpoint data from the node-local storage to a parallel file system;
- compare the transfer parameter to a specified transfer parameter threshold
- determine, based on the comparison of the transfer parameter to the specified transfer parameter threshold, whether to transfer the checkpoint data from the node-local storage to the parallel file system, wherein the transfer parameter includes input/output bandwidth or a percentage disk idle time;
- and
- where the transfer parameter includes the input/output bandwidth, based on a determination that the input/output bandwidth is less than the specified input/output bandwidth threshold, cause the transfer of the checkpoint data from the node-local storage to the parallel file system;
- or
- where the transfer parameter includes the percentage disk idle time, based on a determination that the percentage disk idle time is less than the specified percentage disk idle time threshold, cause the transfer of the checkpoint data from the node-local storage to the parallel file system.
A computer implemented method comprising:
- comparing a storage parameter associated with storage of checkpoint data to a specified storage parameter threshold;
- determining, based on the comparison, whether to store the checkpoint data in a higher order memory tier or in a lower order memory tier, wherein the lower order memory tier is to be used upon occurrence of spillover associated with the higher order memory tier;
- based on a determination that a value of the storage parameter is greater than or equal to a value of the specified storage parameter threshold, storing the checkpoint data in the lower order memory tier;
- and
- determining, based on a comparison of a transfer parameter to a specified transfer parameter threshold, whether to transfer the checkpoint data from a node-local storage of the lower order memory tier to a parallel file system of the lower order memory tier.
A non-transitory computer readable medium having stored thereon machine readable instructions, the machine readable instructions, when executed, cause a processor to:
- ascertain a storage parameter associated with storage of checkpoint data;
- compare the storage parameter to a specified storage parameter threshold;
- and
- determine, based on the comparison of the storage parameter to the specified storage parameter threshold, whether to store the checkpoint data in a higher order memory tier or in a lower order memory tier, wherein the lower order memory tier is to be used upon occurrence of spillover associated with the higher order memory tier.
Dependencies are inferred from the claim text.