Abstract: This disclosure relates to compositions and conjugates comprising an interleukin, e.g., IL-15 and a polypeptide that specifically binds TGF-beta. Typically the polypeptide that specifically binds TGF-beta is a type II TGF-beta receptor. In certain embodiments, the type II TGF-beta receptor is a human isoform, fragment or variant thereof. Uses for treating or preventing cancer and infectious diseases are contemplated.

Abstract: WORM safeguards are provided in a magnetic disk drive in which the disks are pre-magnetized or pre-recorded such that all magnetic domains on each recording surface are completely aligned pointing in the same direction. A special write head which can magnetically write only in the direction opposite to the pre-magnetization orientation is also provided.

Abstract: For non-bursty data writes, data is written to flash memory of a hard disk drive for subsequent de-staging to disk, whereas for bursty writes data is written directly to disk.

Abstract: A mobile computing hard disk drive has both a flash memory device and a DRAM device, with the HDD controller managing data storage between disk, DRAM, and flash both when write requests arrive and when the HDD is idle to optimize flash memory device life and system performance.

Abstract: In a disk drive that uses large block sizes (e.g., 4 KB) for storing data and that responds to read and write requests from a client that uses small block sizes (e.g., 512 bytes), at least the starting and ending 4K blocks of read data are cached. Since much disk data that is the subject of a write request is first read, upon a subsequent write request the drive controller determines whether the starting and ending blocks are in cache and if so, writes new data to those blocks, calculates a full ECC for them, and then calculates ECC for intervening blocks and writes new data to the intervening blocks. If both starting and ending blocks are not in cache the drive controller executes either a high data integrity routine or a high performance routine as chosen by the user.

Abstract: A RAID subsystem with a plurality of hard disk drives has each disk controller autonomously executing data scrubbing on its disks instead of requiring the RAID controller to execute data scrubbing. The data scrubbing may be interrupted upon an I/O request from the RAID controller, or it may continue if a guaranteed scrub rate is not being achieved. Data is stored in large bands of concentric data tracks, and a band may be preferentially selected for scrubbing based on frequency of reading the band, or recency of writing to the band.

Abstract: When writing to a hard disk drive, an effective track skew is used which differs from the track skew used when reading from the disk. During reads from the disk, the disk controller uses a traditional logical track layout and logical block address (LBA) numbering having a track skew value related to read settle time. A different effective track skew value, optimized for write settle time, is used, however, when writing to the disk. Logical blocks are written to the disk out-of-sequence in order to accommodate this different effective track skew value.

Abstract: Upon receiving a read command, a disk drive moves a read head to target data and reads the data into a read buffer. In an action called “prefetching”, the drive continues to read nearby data into the read buffer which doubles as a data cache. When another I/O command is present and must be serviced, prefetching is preempted thereby reducing the data read into the cache. Moving the head from the current I/O command to the next I/O command creates a delay comprising two components: seek time and rational latency. Based on the relative values of these components, a time period, less than the entire delay period, is calculated in which prefetching will continue. By continuing prefetching instead of preempting it, the likelihood of cache hits is increased because more data is available in the read buffer. Furthermore, by performing prefetching during part of the otherwise unused delay period, no performance penalty is introduced.

Abstract: A method and apparatus are disclosed which minimize the additional read and write loading for each remaining DASD in a DASD array occasioned by a single point of failure where subsets of DASDs may accessed asynchronously (RAID 4 or RAID 5). In this method, parity groups are written across the array in a pattern forming a balanced incomplete block design or the equivalent. In the event of failure of a DASD in the array, then any additional loading resulting from the failure is uniformly spread among the remaining DASDs.