"Thrash" the OSDs by randomly marking them out/down (and then back in) until the task is ended. This loops, and every op_delay seconds it randomly chooses to add or remove an OSD (even odds) unless there are fewer than min_out OSDs out of the cluster, or more than min_in OSDs in
(ctx, config)
| 11 | |
| 12 | @contextlib.contextmanager |
| 13 | def task(ctx, config): |
| 14 | """ |
| 15 | "Thrash" the OSDs by randomly marking them out/down (and then back |
| 16 | in) until the task is ended. This loops, and every op_delay |
| 17 | seconds it randomly chooses to add or remove an OSD (even odds) |
| 18 | unless there are fewer than min_out OSDs out of the cluster, or |
| 19 | more than min_in OSDs in the cluster. |
| 20 | |
| 21 | All commands are run on mon0 and it stops when __exit__ is called. |
| 22 | |
| 23 | The config is optional, and is a dict containing some or all of: |
| 24 | |
| 25 | cluster: (default 'ceph') the name of the cluster to thrash |
| 26 | |
| 27 | min_in: (default 4) the minimum number of OSDs to keep in the |
| 28 | cluster |
| 29 | |
| 30 | min_out: (default 0) the minimum number of OSDs to keep out of the |
| 31 | cluster |
| 32 | |
| 33 | op_delay: (5) the length of time to sleep between changing an |
| 34 | OSD's status |
| 35 | |
| 36 | min_dead: (0) minimum number of osds to leave down/dead. |
| 37 | |
| 38 | max_dead: (0) maximum number of osds to leave down/dead before waiting |
| 39 | for clean. This should probably be num_replicas - 1. |
| 40 | |
| 41 | clean_interval: (60) the approximate length of time to loop before |
| 42 | waiting until the cluster goes clean. (In reality this is used |
| 43 | to probabilistically choose when to wait, and the method used |
| 44 | makes it closer to -- but not identical to -- the half-life.) |
| 45 | |
| 46 | scrub_interval: (-1) the approximate length of time to loop before |
| 47 | waiting until a scrub is performed while cleaning. (In reality |
| 48 | this is used to probabilistically choose when to wait, and it |
| 49 | only applies to the cases where cleaning is being performed). |
| 50 | -1 is used to indicate that no scrubbing will be done. |
| 51 | |
| 52 | chance_down: (0.4) the probability that the thrasher will mark an |
| 53 | OSD down rather than marking it out. (The thrasher will not |
| 54 | consider that OSD out of the cluster, since presently an OSD |
| 55 | wrongly marked down will mark itself back up again.) This value |
| 56 | can be either an integer (eg, 75) or a float probability (eg |
| 57 | 0.75). |
| 58 | |
| 59 | chance_test_min_size: (0) chance to run test_pool_min_size, |
| 60 | which: |
| 61 | - kills all but one osd |
| 62 | - waits |
| 63 | - kills that osd |
| 64 | - revives all other osds |
| 65 | - verifies that the osds fully recover |
| 66 | |
| 67 | test_min_size_duration: (1800) the number of seconds for |
| 68 | test_pool_min_size to last. |
| 69 | |
| 70 | timeout: (360) the number of seconds to wait for the cluster |
nothing calls this directly
no test coverage detected