Lsi Corporation

Avago Technologies General IP (singapore) Pte. Ltd.

Varadaraj S Talamacki

Vinu Velayudhan

Senthil Thangaraj

Sumant Kumar Patro

Methods and structure for predictively caching data to service host write requests. One embodiment includes a Redundant Array of Independent Disks (RAID) storage controller able to manage a RAID volume implemented on a plurality of storage devices. The storage controller includes a memory able to store a queue of write requests that are directed to the RAID volume, and an Input/Output (I/O) processor. The I/O processor is able to detect a stream of write requests in the queue that are directed to a sequence of Logical Block Addresses (LBAs) at the RAID volume, to predict, based on the stream, LBAs for new write requests, to cache data for the predicted LBAs from the storage devices to the storage controller, to receive the new write requests, and to utilize the cached data from the predicted LBAs to generate parity data for the new write requests.

Scaling performance for raid storage controllers by predictively caching data for host write requests

Disclosed is a virtual machine system where hardware timer interrupts are processed by a first virtual machine. The first virtual machine writes a timer value to a shared memory location while processing the hardware timer interrupt. The timer value may be based on a kernel timing parameter maintained by the operating system of the first virtual machine. A second virtual machine may read the shared timer value from the shared memory location in order to time inter-virtual machine processes such as I/O processing and I/O requests.

Inter-virtual machine time profiling of I/O transactions

Disclosed is a system with multiple virtual machines passing I/O requests via a shared memory space. A flag in shared memory is set to a first state in response to a first hypervisor I/O interrupt to indicate that an I/O processing routine is active (running). I/O requests are retrieved from an I/O queue in the shared memory by the I/O processing routine. Based on an indicator that there are no I/O requests remaining in said I/O queue, the shared flag is set to a second state to indicate that the I/O processing routine is deactivated (sleeping). In response to said shared flag being in the second state, when new I/O requests are going to be made, a second hypervisor I/O interrupt is generated. In response to said shared flag being in said first state, I/O requests are inserted into the I/O queue without generating a second hypervisor I/O interrupt.

Growing community of inventors

Fremont, CA, United States of America

Varadaraj S Talamacki