Josef Bacik | 3407ef5 | 2011-03-24 13:54:24 +0000 | [diff] [blame] | 1 | dm-flakey |
| 2 | ========= |
| 3 | |
Mike Snitzer | b26f5e3 | 2011-08-02 12:32:05 +0100 | [diff] [blame] | 4 | This target is the same as the linear target except that it exhibits |
| 5 | unreliable behaviour periodically. It's been found useful in simulating |
| 6 | failing devices for testing purposes. |
Josef Bacik | 3407ef5 | 2011-03-24 13:54:24 +0000 | [diff] [blame] | 7 | |
| 8 | Starting from the time the table is loaded, the device is available for |
Mike Snitzer | b26f5e3 | 2011-08-02 12:32:05 +0100 | [diff] [blame] | 9 | <up interval> seconds, then exhibits unreliable behaviour for <down |
| 10 | interval> seconds, and then this cycle repeats. |
Josef Bacik | 3407ef5 | 2011-03-24 13:54:24 +0000 | [diff] [blame] | 11 | |
Mike Snitzer | b26f5e3 | 2011-08-02 12:32:05 +0100 | [diff] [blame] | 12 | Also, consider using this in combination with the dm-delay target too, |
| 13 | which can delay reads and writes and/or send them to different |
| 14 | underlying devices. |
| 15 | |
| 16 | Table parameters |
| 17 | ---------------- |
| 18 | <dev path> <offset> <up interval> <down interval> \ |
| 19 | [<num_features> [<feature arguments>]] |
| 20 | |
| 21 | Mandatory parameters: |
Josef Bacik | 3407ef5 | 2011-03-24 13:54:24 +0000 | [diff] [blame] | 22 | <dev path>: Full pathname to the underlying block-device, or a |
| 23 | "major:minor" device-number. |
| 24 | <offset>: Starting sector within the device. |
| 25 | <up interval>: Number of seconds device is available. |
| 26 | <down interval>: Number of seconds device returns errors. |
Mike Snitzer | b26f5e3 | 2011-08-02 12:32:05 +0100 | [diff] [blame] | 27 | |
| 28 | Optional feature parameters: |
| 29 | If no feature parameters are present, during the periods of |
| 30 | unreliability, all I/O returns errors. |
| 31 | |
| 32 | drop_writes: |
| 33 | All write I/O is silently ignored. |
| 34 | Read I/O is handled correctly. |
Mike Snitzer | a399879 | 2011-08-02 12:32:06 +0100 | [diff] [blame] | 35 | |
| 36 | corrupt_bio_byte <Nth_byte> <direction> <value> <flags>: |
| 37 | During <down interval>, replace <Nth_byte> of the data of |
| 38 | each matching bio with <value>. |
| 39 | |
| 40 | <Nth_byte>: The offset of the byte to replace. |
| 41 | Counting starts at 1, to replace the first byte. |
| 42 | <direction>: Either 'r' to corrupt reads or 'w' to corrupt writes. |
| 43 | 'w' is incompatible with drop_writes. |
| 44 | <value>: The value (from 0-255) to write. |
| 45 | <flags>: Perform the replacement only if bio->bi_rw has all the |
| 46 | selected flags set. |
| 47 | |
| 48 | Examples: |
| 49 | corrupt_bio_byte 32 r 1 0 |
| 50 | - replaces the 32nd byte of READ bios with the value 1 |
| 51 | |
| 52 | corrupt_bio_byte 224 w 0 32 |
| 53 | - replaces the 224th byte of REQ_META (=32) bios with the value 0 |