Abstract
Future main memory will likely include Non-Volatile Memory. Non-Volatile Main Memory (NVMM) provides an opportunity to rethink checkpointing strategies for providing failure safety to applications. While there are many checkpointing and logging schemes in the literature, their use must be revisited as they incur high execution time overheads as well as a large number of additional writes to NVMM, which may significantly impact write endurance. In this article, we propose a novel recompute-based failure safety approach and demonstrate its applicability to loop-based code. Rather than keeping a fully consistent logging state, we only log enough state to enable recomputation. Upon a failure, our approach recovers to a consistent state by determining which parts of the computation were not completed and recomputing them. Effectively, our approach removes the need to keep checkpoints or logs, thus reducing execution time overheads and improving NVMM write endurance 8 at the expense of more complex recovery. We compare our new approach against logging and checkpointing on five scientific workloads, including tiled matrix multiplication, on a computer system model that was built on gem5 and supports Intel PMEM instruction extensions. For tiled matrix multiplication, our recompute approach incurs an execution time overhead of only 5%, in contrast to 8% overhead with logging and 207% overhead with checkpointing. Furthermore, recompute only adds 7% additional NVMM writes, compared to 111% with logging and 330% with checkpointing. We also conduct experiments on real hardware, allowing us to run our workloads to completion while varying the number of threads used for computation. These experiments substantiate our simulation-based observations and provide a sensitivity study and performance comparison between the Recompute Scheme and Naive Checkpointing.
| Original language | English |
|---|---|
| Article number | 18 |
| Journal | ACM Transactions on Architecture and Code Optimization |
| Volume | 16 |
| Issue number | 2 |
| DOIs | |
| Publication status | Published - 2019 May |
Fingerprint
Keywords
- Computer architecture
- Emerging memory technologies
- Memory systems
ASJC Scopus subject areas
- Software
- Information Systems
- Hardware and Architecture
Cite this
Efficient checkpointing with recompute scheme for non-volatile main memory. / Alshboul, Mohammad; Elnawawy, Hussein; Elkhouly, Reem; Kimura, Keiji; Tuck, James; Solihin, Yan.
In: ACM Transactions on Architecture and Code Optimization, Vol. 16, No. 2, 18, 05.2019.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - Efficient checkpointing with recompute scheme for non-volatile main memory
AU - Alshboul, Mohammad
AU - Elnawawy, Hussein
AU - Elkhouly, Reem
AU - Kimura, Keiji
AU - Tuck, James
AU - Solihin, Yan
PY - 2019/5
Y1 - 2019/5
N2 - Future main memory will likely include Non-Volatile Memory. Non-Volatile Main Memory (NVMM) provides an opportunity to rethink checkpointing strategies for providing failure safety to applications. While there are many checkpointing and logging schemes in the literature, their use must be revisited as they incur high execution time overheads as well as a large number of additional writes to NVMM, which may significantly impact write endurance. In this article, we propose a novel recompute-based failure safety approach and demonstrate its applicability to loop-based code. Rather than keeping a fully consistent logging state, we only log enough state to enable recomputation. Upon a failure, our approach recovers to a consistent state by determining which parts of the computation were not completed and recomputing them. Effectively, our approach removes the need to keep checkpoints or logs, thus reducing execution time overheads and improving NVMM write endurance 8 at the expense of more complex recovery. We compare our new approach against logging and checkpointing on five scientific workloads, including tiled matrix multiplication, on a computer system model that was built on gem5 and supports Intel PMEM instruction extensions. For tiled matrix multiplication, our recompute approach incurs an execution time overhead of only 5%, in contrast to 8% overhead with logging and 207% overhead with checkpointing. Furthermore, recompute only adds 7% additional NVMM writes, compared to 111% with logging and 330% with checkpointing. We also conduct experiments on real hardware, allowing us to run our workloads to completion while varying the number of threads used for computation. These experiments substantiate our simulation-based observations and provide a sensitivity study and performance comparison between the Recompute Scheme and Naive Checkpointing.
AB - Future main memory will likely include Non-Volatile Memory. Non-Volatile Main Memory (NVMM) provides an opportunity to rethink checkpointing strategies for providing failure safety to applications. While there are many checkpointing and logging schemes in the literature, their use must be revisited as they incur high execution time overheads as well as a large number of additional writes to NVMM, which may significantly impact write endurance. In this article, we propose a novel recompute-based failure safety approach and demonstrate its applicability to loop-based code. Rather than keeping a fully consistent logging state, we only log enough state to enable recomputation. Upon a failure, our approach recovers to a consistent state by determining which parts of the computation were not completed and recomputing them. Effectively, our approach removes the need to keep checkpoints or logs, thus reducing execution time overheads and improving NVMM write endurance 8 at the expense of more complex recovery. We compare our new approach against logging and checkpointing on five scientific workloads, including tiled matrix multiplication, on a computer system model that was built on gem5 and supports Intel PMEM instruction extensions. For tiled matrix multiplication, our recompute approach incurs an execution time overhead of only 5%, in contrast to 8% overhead with logging and 207% overhead with checkpointing. Furthermore, recompute only adds 7% additional NVMM writes, compared to 111% with logging and 330% with checkpointing. We also conduct experiments on real hardware, allowing us to run our workloads to completion while varying the number of threads used for computation. These experiments substantiate our simulation-based observations and provide a sensitivity study and performance comparison between the Recompute Scheme and Naive Checkpointing.
KW - Computer architecture
KW - Emerging memory technologies
KW - Memory systems
UR - http://www.scopus.com/inward/record.url?scp=85069190431&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85069190431&partnerID=8YFLogxK
U2 - 10.1145/3323091
DO - 10.1145/3323091
M3 - Article
AN - SCOPUS:85069190431
VL - 16
JO - Transactions on Architecture and Code Optimization
JF - Transactions on Architecture and Code Optimization
SN - 1544-3566
IS - 2
M1 - 18
ER -