Conversation
89ba2c6 to
dbf1a66
Compare
| cache.reset_counters() | ||
|
|
||
| with TestRun.step(f"Fill exported object with dirty data"): | ||
| with TestRun.step(f"Prepare I/O to fill exported object with dirty data"): |
There was a problem hiding this comment.
unnecessary f-string
how is I/O prepared here? (preparing config for I/O tool is not a real test step)
| dirty_4k = cache.get_statistics().usage_stats.dirty.get_value(Unit.Blocks4096) | ||
| fio.verification_with_pattern("0xdeadbeef") | ||
|
|
||
| with TestRun.step("Run fio"): |
There was a problem hiding this comment.
I didn't see this step before writing my comment for the previous file ;)
| with TestRun.step("Prepare I/O to separate disk test file"): | ||
| fio = ( | ||
| Fio() | ||
| .create_command() | ||
| .io_engine(IoEngine.libaio) | ||
| .read_write(ReadWrite.write) | ||
| .block_size(bs) | ||
| .block_size(Size(2, Unit.MebiByte)) | ||
| .direct() | ||
| .io_depth(256) | ||
| .target(core) | ||
| .size(file_size) | ||
| ) | ||
| fio.default_run_time = timedelta(hours=4) # timeout for non-time-based fio | ||
|
|
||
| with TestRun.step("Run I/O"): |
| core_disk = TestRun.disks["core"] | ||
|
|
||
| cache_disk.create_partitions([Size(100, Unit.MebiByte)]) | ||
| core_disk.create_partitions([Size(5, Unit.GibiByte)] * 2) |
There was a problem hiding this comment.
2 is a small number of partitions in a test the randomises the order of adding them
|
|
||
| for _ in TestRun.iteration(range(0, iterations_number)): | ||
| with TestRun.step("Detach all core devices"): | ||
| for core in cores: |
There was a problem hiding this comment.
I think detach order should be also randomized
| """ | ||
| title: Test for triple level cache. | ||
| description: | | ||
| Test if cache can be used as core of other cache up to three levels |
There was a problem hiding this comment.
this is incorrect - cache can never be used as a core
| with TestRun.step("Create file on exported object"): | ||
| md5_sums = create_files_with_md5sums(mount_point, 100) | ||
|
|
||
| with TestRun.step("Check md5 sums between original file and exported object"): | ||
| compare_md5sums(md5_sums, mount_point, copy_to_tmp=True) |
There was a problem hiding this comment.
confusing steps - if you create the file on the exported object, then if is the original file, so the last step implies comparing the test file to itself ;)
Signed-off-by: Kamil Gierszewski <kamil.gierszewski@huawei.com>
…hronous read of the device Signed-off-by: Kamil Gierszewski <kamil.gierszewski@huawei.com>
Signed-off-by: Kamil Gierszewski <kamil.gierszewski@huawei.com>
Signed-off-by: Kamil Gierszewski <kamil.gierszewski@huawei.com>
Signed-off-by: Kamil Gierszewski <kamil.gierszewski@huawei.com>
dbd184f to
237e25d
Compare
No description provided.