International Business Machines Corporation

David Alan Hepkin

Bret Ronald Olszewski

Thomas Stanley Mathews

Naresh Nayar

Randal Craig Swanberg

Bruce Gerard Mealey

Andrew Dunshea

Vaijayanthimala K Anand

Stuart Zachary Jacobs

Satya Prakash Sharma

Jos Manuel Accapadi

Mark D Rogers

Li Li

Rajeev Kumar Mishra

Perinkulam I Ganesh

Mysore Sathyanarayana Srinivas

Wade Byron Ouren

Grover Herbert Neuman

Matthew David Fleming

Saurabh Nath Sharma

Saravanan Devendra

Rajalakshmi SrinivasaRaghavan

Vinit Jain

Richard Louis Arndt

Matthew Albert Huras

Kyle Alan Lucke

Dirk Michel

Peter Joseph Heyrman

Christian Marcelo Garcia-Arellano

Joefon Jann

Kenneth Charles Vossen

Sujatha Kashyap

Jorge Rafael Nogueras

Aamer U Sachedina

Andrew T Koch

Carol B Hernandez

Ramanjaneya Sarma Burugula

Sergio Reyes

Stephen Bailey Peckham

Andrew David Buckler

Randall Craig Swanberg

Memory page de-duplication in a computer system that includes a plurality of virtual machine partitions managed by a hypervisor, where each virtual machine is assigned a different dedicated memory partition, may include: identifying, by the hypervisor, a plurality of identical memory pages in memory of one or more dedicated memory partitions; assigning, by the hypervisor, one of the identical memory pages as a master page; mapping, for each virtual machine having an identical memory page, each of the identical memory pages to the master page; and directing, by the hypervisor, reads of the memory page to the master page.

Memory page de-duplication in a computer system that includes a plurality of virtual machines

A system, and computer usable program product for expanding memory size are provided in the illustrative embodiments. A desired size of an expanded memory and a first information about a workload in the data processing system are received. A size of a compressed memory pool to use with the memory to make the desired size of the expanded memory available is computed. A representation of the memory is configured, the representation of the memory appearing to be of a size larger than the size of the memory, the representation of the memory being the expanded memory, and the size of the representation being the size of the expanded memory. The expanded memory is made available such that the memory in the data processing system is usable by addressing the expanded memory.

Expanding memory size

A method, system, and computer usable program product for selective memory donation in a virtual real memory environment are provided in the illustrative embodiments. A virtual machine receives a request for memory donation. A component of the virtual machine determines whether a portion of a memory space being used for file caching exceeds a threshold. The determining forms a threshold determination, and the portion of the memory space being used for file caching forms a file cache. If the threshold determination is false, the component ignores the request. If the threshold determination is true, a component of the virtual machine releases a part of the file cache that exceeds the threshold. The part of the file cache forms a released file cache. In response to the request, the virtual machine makes the released file cache available to a requester of the request.

Selective memory donation in virtual real memory environment

Two or more processors that each provides a specified thread to access a shared resource that can only be accessed by one thread at a given time. A locking mechanism enables one of the threads to access the shared resource while other threads are retained in a waiting queue. Responsive to an additional thread that is not one of the specified threads being provided access the shared resource during an identified time period, and responsive to a first criterion an a second criterion being met, the additional thread accesses the shared resource before the other threads in the waiting queue.

Lock mechanism to reduce waiting of threads to access a shared resource by selectively granting access to a thread before an enqueued highest priority thread

A method of dynamically reallocating memory affinity in a virtual machine after migrating the virtual machine from a source computer system to a destination computer system migrates processor states and resources used by the virtual machine from the source computer system to the destination computer system. The method maps memory of the virtual machine to processor nodes of the destination computer system. The method deletes memory mappings in processor hardware, such as translation lookaside buffers and effective-to-real address tables, for the virtual machine on the destination computer system. The method starts the virtual machine on the destination computer system in virtual real memory mode. A hypervisor running on the destination computer system receives a page fault and virtual address of a page for said virtual machine from a processor of the destination computer system and determines if the page is in local memory of the processor. If the hypervisor determines the page to be in the local memory of the processor, the hypervisor returning a physical address mapping for the page to the processor. If the hypervisor determines the page not to be in the local memory of the processor, the hypervisor moves the page to local memory of the processor and returns a physical address mapping for said page to the processor.

Dynamic memory affinity reallocation after partition migration

A method includes atomically reading a next field of a current element of the linked list to determine a first value that encodes a first pointer to the first element and a first indication of an owner of the first element. The first indication of the owner is stored in a first of a plurality of multi-field reservation data structures. The operation includes determining whether the next field of the current element still indicates the first value. The operation includes reading the first element of the linked list via the first pointer if the next field of the current element still indicates the first value. If the next field of the current element indicates a current value different than the first value, the first indication of the owner is removed from the first multi-field reservation data structure, and storing and determining with the second value is repeated.

Maintenance and access of a linked list

A system, and computer usable program product for modeling memory compression are provided in the illustrative embodiments. A subset of candidate pages is received. The subset of candidate pages is a subset of a set of candidate pages used in executing a workload in a data processing system. A candidate page is compressible uncompressed data in a memory associated with the data processing system. The subset of candidate pages is compressed in a scratch space. A compressibility of the workload is computed based on the compression of the subset of candidate pages. Page reference information of the subset of candidate pages is received. A memory reference rate of the workload is determined A recommendation is presented about a memory compression model for the workload in the data processing system.

Modeling memory compression

A method for modeling memory compression is provided in the illustrative embodiments. A subset of candidate pages is received. The subset of candidate pages is a subset of a set of candidate pages used in executing a workload in a data processing system. A candidate page is compressible uncompressed data in a memory associated with the data processing system. The subset of candidate pages is compressed in a scratch space. A compressibility of the workload is computed based on the compression of the subset of candidate pages. Page reference information of the subset of candidate pages is received. A memory reference rate of the workload is determined. A recommendation is presented about a memory compression model for the workload in the data processing system.

A method, computer program product, and system for detecting and fencing off a failed entity instance so that failover time in the computing system is reduced. Upon detection of a failed entity, for example a failed process, a signal handler invokes an operating system interface to fence off the failed instance from modifying its persistent state, so that the entity may be re-instantiated prior to termination of the failed instance. This approach reduces failover time and eliminates split-brain problems without compromising access to the failed instance to obtain diagnostic information, core dumps, and the like.

Failure detection and fencing in a computing system

An approach is provided that uses a hypervisor to allocate a shared memory pool amongst a set of partitions (e.g., guest operating systems) being managed by the hypervisor. The hypervisor retrieves memory related metrics from shared data structures stored in a memory, with each of the shared data structures corresponding to a different one of the partitions. The memory related metrics correspond to a usage of the shared memory pool allocated to the corresponding partition. The hypervisor identifies a memory stress associated with each of the partitions with this identification based in part on the memory related metrics retrieved from the shared data structures. The hypervisor then reallocates the shared memory pool amongst the plurality of partitions based on the identified memory stress of the plurality of partitions.

Characterizing multiple resource utilization using a relationship model to optimize memory utilization in a virtual machine environment

An approach is provided that retrieves a time spent value corresponding to a selected partition that is selected from a group of partitions included in a virtualized environment running on a computer system. The virtualized environment is provided by a Hypervisor. The time spent value corresponds to an amount of time the selected partition has spent processing interrupts. A number of virtual CPUs have been assigned to the selected partition. The time spent value (e.g., a percentage of the time that the selected partition spends processing interrupts) is compared to one or more interrupt threshold values. If the comparison reveals that the time that the partition is spending processing interrupts exceeds a threshold, then the number of virtual CPUs assigned to the selected partition is increased.

System and method for optimizing interrupt processing in virtualized environments

Disclosed is a computer implemented method and computer program product to prioritize paging-in pages in a remote paging device. An arrival machine receives checkpoint data from a departure machine. The arrival machine restarts at least one process corresponding to the checkpoint data. The arrival machine determines whether a page associated with the process is pinned. The arrival machine associates the page to the remote paging device, responsive to a determination that the page is pinned. The arrival machine touches the page.

Mapped offsets preset ahead of process migration

In one embodiment a method for migrating a workload from one processing resource to a second processing resource of a computing platform is disclosed. The method can include a command to migrate a workload that is processing and the process can be interrupted and some memory processes can be frozen in response to the migration command. An index table can be created that identifies memory locations that determined where the process was when it is interrupted. Table data, pinned page data, and non-private process data can be sent to the second processing resource. Contained in this data can be restart type data. The second resource or target resource can utilize this data to restart the process without the requirement of bulk data transfers providing an efficient migration process. Other embodiments are also disclosed.

Workload migration using on demand remote paging

Growing community of inventors

Austin, TX, United States of America

David Alan Hepkin