Vmware, Inc.

Leland Stanford Junior University

James Eugene Chow

Tal Garfinkel

Peter Chen

Constantine P Sapuntzakis

Monica Sin-Ling Lam

David James Brumley

Mendel Rosenblum

Ramesh U V Chandra

Dominic Lucchetti

Nickolai Borisovich Zeldovich

Boris Weissman

Jeffrey W Sheldon

Ganesh Venkitachalam

Vyacheslav Vladimirovich Malyugin

Min Xu

Petr Vandrovec

Radu Rugina

Dmitry Grinberg

Kalaiselvi Sengottuvel

Jiajun Cao

Srihari Venkatesan

Alexander Dean Gotsis

Andrew Biggadike

Techniques for implementing IOMMU-based DMA tracking for enabling live migration of VMs that use passthrough physical devices are provided. In one set of embodiments, these techniques leverage an IOMMU feature known as dirty bit tracking which is available in most, if not all, modern IOMMU implementations. The use of this feature allows for the tracking of passthrough DMA in a manner that is device/vendor/driver agnostic, resulting in a solution that is universally applicable to all passthrough physical devices.

IOMMU-based direct memory access (DMA) tracking for enabling live migration of virtual machines (VMS) using passthrough physical devices

Dynamic program analysis is decoupled from execution in virtual computer environments so that program analysis can be performed on a running computer program without affecting or perturbing the workload of the system on which the program is executing. Decoupled dynamic program analysis is enabled by separating execution and analysis into two tasks: (1) recording, where system execution is recorded with minimal interference, and (2) analysis, where the execution is replayed and analyzed.

Decoupling dynamic program analysis from execution in virtual environments

The method of taking checkpoints of a virtual computing instance running in a first host computer at varying time intervals includes queuing outgoing network data packets of the virtual computing instance during execution of the virtual computing instance, triggering a process to take the checkpoint, the process including generating a checkpoint information packet and transmitting the checkpoint information packet to a second host computer, and upon receiving acknowledgement from the second host computer of receipt of the checkpoint information packet, releasing the outgoing network data packets of the virtual computing instance that have been queued and not released. The triggering time for the process to take the checkpoint is dependent on characteristics of workload executed by the virtual computing instance.

Method and system for taking checkpoints at varying intervals

During checkpointing of a virtual computing instance running in a first host computer, memory pages that have been modified during a checkpoint time interval are lazily transmitted from the first host computer to a second, backup, host computer. The lazy transmission technique includes the steps of stunning the virtual computing instance at the end of the time interval, and after said stunning, (i) detecting at the first host computer the memory pages that have been modified during the time interval, and (ii) copying one or more of the modified memory pages to a memory region in the first host computer. Then, after the virtual computing instance resumes execution, i.e., while the virtual computing instance is running in the first host computer, the copied memory pages are transmitted from the memory region in the first host computer to the second host computer.

Lazy transmission of memory pages during checkpointing

Execution behavior for processes within a virtual machine is recorded for subsequent replay. The execution behavior comprises a detailed, low-level recording of state changes for processes within the virtual machine. The low-level recording is processed via replay to produce a sliced recording that conforms to time, abstraction, and security requirements for a specific replay scenario. Multiple stages of replay may be arbitrarily stacked to generate different crosscut versions of a common low-level recording.

System and method of manipulating virtual machine recordings for high-level execution and replay

A virtual machine system decouples dynamic program analysis from program execution. Program analysis is decoupled from program execution through the use of a virtual machine to record program execution and an analysis platform to replay and analyze the program execution. Optimization techniques are applied to prevent the analysis platform from falling too far behind the program execution platform during replay.

Accelerating replayed program execution to support decoupled program analysis

The execution behavior of a selected application is recorded for subsequent replay. During recording, only those portions of memory that are accessed by the selected application are stored. As a result, the amount of data that is stored during the recording session is reduced and data that is not necessary for replaying the selected application, which may include possible sensitive and personal information, are not stored.

Method and system for recording a selected computer process for subsequent replay

Dynamic program analysis is decoupled from execution in virtual computer environments so that program analysis can be performed on a running computer program without affecting or perturbing the workload of the system on which the program is executing. Decoupled dynamic program analysis is enabled by separating execution and analysis into two tasks: (1) recording, where system execution is recorded with minimal interference, and (2) analysis, where the execution is replayed and analyzed. Recording and analysis are carried out on heterogeneous systems so that they can be separately optimized.

Decoupling dynamic program analysis from execution across heterogeneous systems

Replay-time-only functionalities in a computer program are executed only during replay in a virtual machine and are skipped outside of replay. If a replay-time-only functionality is detected during the replay of a program execution in a virtual machine, the replay may be paused and the virtual machine state may be saved. The replay-time-only core functionality is executed. When this execution is complete, a prior state of the virtual machine may be restored and the replay may be resumed.

Replay time only functionalities in a virtual machine

Various approaches for virtual appliance management are described. In one approach a virtual appliance repository stores one or more virtual appliances and is coupled to the host computer via a network. A storage device stores a transceiver program capable when executed on said host computer of requesting and receiving the virtual appliances, and generating for each received virtual appliance a respective local copy on the host computer of each received virtual appliance. The local copy is private to the host computer. The transceiver program further binds the virtual appliances to the host computer and obtains user data relevant to the virtual appliances. The transceiver program runs each of the virtual appliances from the respective private local copies on the host computer.

Virtual appliance management

A cache-based system management architecture named 'the Collective' automates system management and supports mobile computing. The Collective manages systems by manipulating virtual appliances, each encapsulating a system state. A virtual appliance (VA) may comprise a virtual network of VAs. Distributed repositories separately storing the VAs and user data. Publishers create and maintain VAs. A user plugs a bootable portable storage device implementing a VA transceiver (VAT) into a host computer. The VAT, comprising a Manager, a Cache Client, and a virtual machine monitor (VMM), authenticates the user, retrieves a copy of the latest version of a subscribed VA from the appropriate network repository, obtains user data, and automatically backs up the user data over the network in the background while running the subscribed VA. Multiple VAs can be run at the same time. With the novel caching protocol and cache charging process, the Collective performs well and has numerous advantages.

Growing community of inventors

San Jose, CA, United States of America

James Eugene Chow