| Jens Axboe | cbb5901 | 2009-02-02 13:02:31 +0100 | [diff] [blame] | 1 | Queue sysfs files |
| 2 | ================= |
| 3 | |
| 4 | This text file will detail the queue files that are located in the sysfs tree |
| 5 | for each block device. Note that stacked devices typically do not export |
| 6 | any settings, since their queue merely functions are a remapping target. |
| 7 | These files are the ones found in the /sys/block/xxx/queue/ directory. |
| 8 | |
| 9 | Files denoted with a RO postfix are readonly and the RW postfix means |
| 10 | read-write. |
| 11 | |
| 12 | hw_sector_size (RO) |
| 13 | ------------------- |
| 14 | This is the hardware sector size of the device, in bytes. |
| 15 | |
| 16 | max_hw_sectors_kb (RO) |
| 17 | ---------------------- |
| 18 | This is the maximum number of kilobytes supported in a single data transfer. |
| 19 | |
| 20 | max_sectors_kb (RW) |
| 21 | ------------------- |
| 22 | This is the maximum number of kilobytes that the block layer will allow |
| 23 | for a filesystem request. Must be smaller than or equal to the maximum |
| 24 | size allowed by the hardware. |
| 25 | |
| 26 | nomerges (RW) |
| 27 | ------------- |
| 28 | This enables the user to disable the lookup logic involved with IO merging |
| 29 | requests in the block layer. Merging may still occur through a direct |
| 30 | 1-hit cache, since that comes for (almost) free. The IO scheduler will not |
| 31 | waste cycles doing tree/hash lookups for merges if nomerges is 1. Defaults |
| 32 | to 0, enabling all merges. |
| 33 | |
| 34 | nr_requests (RW) |
| 35 | ---------------- |
| 36 | This controls how many requests may be allocated in the block layer for |
| 37 | read or write requests. Note that the total allocated number may be twice |
| 38 | this amount, since it applies only to reads or writes (not the accumulated |
| 39 | sum). |
| 40 | |
| 41 | read_ahead_kb (RW) |
| 42 | ------------------ |
| 43 | Maximum number of kilobytes to read-ahead for filesystems on this block |
| 44 | device. |
| 45 | |
| 46 | rq_affinity (RW) |
| 47 | ---------------- |
| 48 | If this option is enabled, the block layer will migrate request completions |
| 49 | to the CPU that originally submitted the request. For some workloads |
| 50 | this provides a significant reduction in CPU cycles due to caching effects. |
| 51 | |
| 52 | scheduler (RW) |
| 53 | -------------- |
| 54 | When read, this file will display the current and available IO schedulers |
| 55 | for this block device. The currently active IO scheduler will be enclosed |
| 56 | in [] brackets. Writing an IO scheduler name to this file will switch |
| 57 | control of this block device to that new IO scheduler. Note that writing |
| 58 | an IO scheduler name to this file will attempt to load that IO scheduler |
| 59 | module, if it isn't already present in the system. |
| 60 | |
| 61 | |
| 62 | |
| 63 | Jens Axboe <jens.axboe@oracle.com>, February 2009 |