MPI (Message Passing Interface) is a specification for a standard library for message passing that was defined by the MPI Forum, a broadly based group of parallel computer vendors, library writers, and applications specialists. Multiple implementations of MPI have been developed. In this paper, we describe MPICH, unique among existing implementations in its design goal of combining portability with high performance. We document its portability and performance and describe the architecture by which these features are simultaneously achieved. We also discuss the set of tools that accompany the free distribution of MPICH, which constitute the beginnings of a portable parallel programming environment. A project of this scope inevitably imparts lessons about parallel computing, the specification being followed, the current hardware and software environment for parallel computing, and project management; we describe those we have learned. Finally, we discuss future developments for MPICH, including those necessary to accommodate extensions to the MPI Standard now being contemplated by the MPI Forum.

A high-performance, portable implementation of the MPI message passing interface standard

Portable implementation of the mpi message passing interface standard

This survey covers rollback-recovery techniques that do not require special language constructs. In the first part of the survey we classify rollback-recovery protocols into checkpoint-based and log-based.Checkpoint-based protocols rely solely on checkpointing for system state restoration. Checkpointing can be coordinated, uncoordinated, or communication-induced. Log-based protocols combine checkpointing with logging of nondeterministic events, encoded in tuples called determinants. Depending on how determinants are logged, log-based protocols can be pessimistic, optimistic, or causal. Throughout the survey, we highlight the research issues that are at the core of rollback-recovery and present the solutions that currently address them. We also compare the performance of different rollback-recovery protocols with respect to a series of desirable properties and discuss the issues that arise in the practical implementations of these protocols.

/pdf/a-survey-of-rollback-recovery-protocols-in-message-passing-bnrpate93a.pdf

A survey of rollback-recovery protocols in message-passing systems

Allowing applications to survive hardware failure is an expensive undertaking, which generally involves reengineering software to include complicated recovery logic as well as deploying special-purpose hardware; this represents a severe barrier to improving the dependability of large or legacy applications. We describe the construction of a general and transparent high availability service that allows existing, unmodified software to be protected from the failure of the physical machine on which it runs. Remus provides an extremely high degree of fault tolerance, to the point that a running system can transparently continue execution on an alternate physical host in the face of failure with only seconds of downtime, while completely preserving host state such as active network connections. Our approach encapsulates protected software in a virtual machine, asynchronously propagates changed state to a backup host at frequencies as high as forty times a second, and uses speculative execution to concurrently run the active VM slightly ahead of the replicated system state.

/pdf/remus-high-availability-via-asynchronous-virtual-machine-1alfua9yqt.pdf

Remus: high availability via asynchronous virtual machine replication

This paper shows how to quickly move the state of a running computer across a network, including the state in its disks, memory, CPU registers, and I/O devices. We call this state a capsule. Capsule state is hardware state, so it includes the entire operating system as well as applications and running processes.We have chosen to move x86 computer states because x86 computers are common, cheap, run the software we use, and have tools for migration. Unfortunately, x86 capsules can be large, containing hundreds of megabytes of memory and gigabytes of disk data. We have developed techniques to reduce the amount of data sent over the network: copy-on-write disks track just the updates to capsule disks, "ballooning" zeros unused memory, demand paging fetches only needed blocks, and hashing avoids sending blocks that already exist at the remote end. We demonstrate these optimizations in a prototype system that uses VMware GSX Server virtual, machine monitor to create and run x86 capsules. The system targets networks as slow as 384 kbps.Our experimental results suggest that efficient capsule migration can improve user mobility and system management. Software updates or installations on a set of machines can be accomplished simply by distributing a capsule with the new changes. Assuming the presence of a prior capsule, the amount of traffic incurred is commensurate with the size of the update or installation package itself. Capsule migration makes it possible for machines to start running an application within 20 minutes on a 384 kbps link, without having to first install the application or even the underlying operating system. Furthermore, users' capsules can be migrated during a commute between home and work in even less time.

/pdf/optimizing-the-migration-of-virtual-computers-52xckzpz3v.pdf

Optimizing the migration of virtual computers

Checkpointing of parallel applications can be used as the core technology to provide process migration. Both checkpointing and migration, are an important issue for parallel applications on networks of workstations. The CoCheck environment which we present in this paper introduces a new approach to provide checkpointing and migration for parallel applications. CoCheck sits on top of the message passing library and achieves consistency at a level above the message passing system. It uses an existing single process checkpointer which is available for a wide range of systems. Hence, CoCheck can be easily adapted to both, different message passing systems and new machines.

G. Stellner

Papers

CoCheck: checkpointing and process migration for MPI