An official website of the United States government
Checkpoint/restart (C/R) provides fault-tolerant computing capability, enables long running applications, and provides scheduling flexibility for computing centers to support diverse workloads with different priority. It is therefore vital to get transparent C/R capability working at NERSC. MANA, by Garg et. al., is a transparent checkpointing tool that has been selected due to its MPI-agnostic and network-agnostic approach. However, originally written as a proof-of-concept code, MANA was not ready to use with NERSC's diverse production workloads, which are dominated by MPI and hybrid MPI+OpenMP applications. In this talk, we present ongoing work at NERSC to enable MANA for NERSC's production workloads, including fixing bugs that were exposed by the top applications at NERSC, adding new features to address system changes, evaluating C/R overhead at scale, etc. The lessons learned from making MANA production-ready for HPC applications will be useful for C/R tool developers, supercomputing centers and HPC end-users alike.
more »
« less
Checkpointing SPAdes for Metagenome Assembly: Transparency versus Performance in Production
The SPAdes assembler for metagenome assembly is a long-running application commonly used at the NERSC supercomputing site. However, NERSC, like many other sites, has a 48-hour limit on resource allocations. The solution is to chain together multiple resource allocations in a single run, using checkpoint-restart. This case study provides insights into the "pain points" in applying a well-known checkpointing package (DMTCP: Distributed MultiThreaded CheckPointing) to long-running production workloads of SPAdes. This work has exposed several bugs and limitations of DMTCP, which were fixed to support the large memory and fragmented intermediate files of SPAdes. But perhaps more interesting for other applications, this work reveals a tension between the transparency goals of DMTCP and performance concerns due to an I/O bottleneck during the checkpointing process when supporting large memory and many files. Suggestions are made for overcoming this I/O bottleneck, which provides important "lessons learned" for similar applications.
more »
« less
- Award ID(s):
- 1740218
- PAR ID:
- 10319922
- Date Published:
- Journal Name:
- First Int. Symp. on Checkpointing for Supercomputing (SuperCheck21)
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Writing large amounts of data concurrently to stable storage is a typical I/O pattern of many HPC workflows. This pattern introduces high I/O overheads and results in increased storage space utilization especially for workflows that need to capture the evolution of data structures with high frequency as checkpoints. In this context, many applications, such as graph pattern matching, perform sparse updates to large data structures between checkpoints. For these applications, incremental checkpointing techniques that save only the differences from one checkpoint to another can dramatically reduce the checkpoint sizes, I/O bottlenecks, and storage space utilization. However, such techniques are not without challenges: it is non-trivial to transparently determine what data has changed since a previous checkpoint and assemble the differences in a compact fashion that does not result in excessive metadata. State-of-art data reduction techniques (e.g., compression and de-duplication) have significant limitations when applied to modern HPC applications that leverage GPUs: slow at detecting the differences, generate a large amount of metadata to keep track of the differences, and ignore crucial spatiotemporal checkpoint data redundancy. This paper addresses these challenges by proposing a Merkle tree-based incremental checkpointing method to exploit GPUs’ high memory bandwidth and massive parallelism. Experimental results at scale show a significant reduction of the I/O overhead and space utilization of checkpointing compared with state-of-the-art incremental checkpointing and compression techniques.more » « less
-
De_Vita, R; Espinal, X; Laycock, P; Shadura, O (Ed.)Creating new materials, discovering new drugs, and simulating systems are essential processes for research and innovation and require substantial computational power. While many applications can be split into many smaller independent tasks, some cannot and may take hours or weeks to run to completion. To better manage those longer-running jobs, it would be desirable to stop them at any arbitrary point in time and later continue their computation on another compute resource; this is usually referred to as checkpointing. While some applications can manage checkpointing programmatically, it would be preferable if the batch scheduling system could do that independently. This paper evaluates the feasibility of using CRIU (Checkpoint Restore in Userspace), an open-source tool for the GNU/Linux environments, emphasizing the OSG’s OSPool HTCondor setup. CRIU allows checkpointing the process state into a disk image and can deal with both open files and established network connections seamlessly. Furthermore, it can checkpoint traditional Linux processes and containerized workloads. The functionality seems adequate for many scenarios supported in the OSPool. However, some limitations prevent it from being usable in all circumstances.more » « less
-
The serverless and functions as a service (FaaS) paradigms are currently trending among cloud providers and are now increasingly being applied to the network edge, and to the Internet of Things (IoT) devices. The benefits include reduced latency for communication, less network traffic and increased privacy for data processing. However, there are challenges as IoT devices have limited resources for running multiple simultaneous containerized functions, and also FaaS does not typically support long-running functions. Our implementation utilizes Docker and CRIU for checkpointing and suspending long-running blocking functions. The results show that checkpointing is slightly slower than regular Docker pause, but it saves memory and allows for more long-running functions to be run on an IoT device. Furthermore, the resulting checkpoint files are small, hence they are suitable for live migration and backing up stateful functions, therefore improving availability and reliability of the system.more » « less
-
null (Ed.)Deep learning has been shown as a successful method for various tasks, and its popularity results in numerous open-source deep learning software tools. Deep learning has been applied to a broad spectrum of scientific domains such as cosmology, particle physics, computer vision, fusion, and astrophysics. Scientists have performed a great deal of work to optimize the computational performance of deep learning frameworks. However, the same cannot be said for I/O performance. As deep learning algorithms rely on big-data volume and variety to effectively train neural networks accurately, I/O is a significant bottleneck on large-scale distributed deep learning training. This study aims to provide a detailed investigation of the I/O behavior of various scientific deep learning workloads running on the Theta supercomputer at Argonne Leadership Computing Facility. In this paper, we present DLIO, a novel representative benchmark suite built based on the I/O profiling of the selected workloads. DLIO can be utilized to accurately emulate the I/O behavior of modern scientific deep learning applications. Using DLIO, application developers and system software solution architects can identify potential I/O bottlenecks in their applications and guide optimizations to boost the I/O performance leading to lower training times by up to 6.7x.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
