On Sat, Nov 12, 2011 at 02:57:17PM +0900, MORITA Kazutaka wrote: > At Fri, 11 Nov 2011 23:17:23 +0800, > chenbaozi at gmail.com wrote: > > > > From: CHEN Baozi <chenbaozi.pt at taobao.com> > > > > Signed-off-by: CHEN Baozi <chenbaozi.pt at taobao.com> > > --- > > tests/qemu_io_testcases.py | 159 ++++++++++++++++++++++++++++++++++++++++++++ > > tests/test_qemu_io.py | 84 +++++++++++++++++++++++ > > 2 files changed, 243 insertions(+), 0 deletions(-) > > create mode 100644 tests/qemu_io_testcases.py > > create mode 100644 tests/test_qemu_io.py > > > > diff --git a/tests/qemu_io_testcases.py b/tests/qemu_io_testcases.py > > new file mode 100644 > > index 0000000..6998bbe > > --- /dev/null > > +++ b/tests/qemu_io_testcases.py > > Does this file borrows code from qemu-iotests? If so, I guess you > should add the original copyright. > I borrowed the testcases from qemu-iostests. So what I should do is add the copyright information such as following? " # Copyright (C) 2009 Red Hat, Inc. # # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # # You should have received a copy of the GNU General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. " I'm sorry for not clear on this kind of stuff. > > > @@ -0,0 +1,159 @@ > > +cases_desc = { > > +"001": "Test simple read/write using plain bdrv_read/bdrv_write.", > > +"002": "Test simple read/write using plain bdrv_pread/bdrv_pwrite.", > > +"003": "Test simple read/write using bdrv_aio_readv/bdrv_aio_writev.", > > +"004": "Make sure we can't read and write outside of the image size.", > > +"008": "Test simple asynchronous read/write operations.", > > +"011": "Test for AIO allocation on the same cluster.", > > +"016": "Test I/O after EOF for growable images.", > > +"025": "Resizing images.", > > +} > > + > > +io_cases = { > > +"001":[ > > +("read 0 128M", """read 134217728/134217728 bytes at offset 0 > > +128 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("write -P 0xa 0 128M", """wrote 134217728/134217728 bytes at offset 0 > > +128 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("read -P 0xa 0 128M", """read 134217728/134217728 bytes at offset 0 > > +128 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +], > > +"002":[ > > +("read -p 0 128M", """read 134217728/134217728 bytes at offset 0 > > +128 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("write -pP 0xa 0 128M", """wrote 134217728/134217728 bytes at offset 0 > > +128 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("read -pP 0xa 0 128M", """read 134217728/134217728 bytes at offset 0 > > +128 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("write -pP 0xab 66 42", """wrote 42/42 bytes at offset 66 > > +42.000000 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("read -pP 0xab 66 42", """read 42/42 bytes at offset 66 > > +42.000000 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +], > > +"003":[ > > +("readv 0 128M", """read 134217728/134217728 bytes at offset 0 > > +128 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("writev -P 0xa 0 128M", """wrote 134217728/134217728 bytes at offset 0 > > +128 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("readv -P 0xa 0 128M", """read 134217728/134217728 bytes at offset 0 > > +128 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("writev -P 0xb 67M 8k 8k 8k 8k 8k 8k 8k", > > +"""wrote 57344/57344 bytes at offset 70254592 > > +56 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("readv -P 0xb 67M 8k 8k 8k 8k 8k 8k 8k", > > +"""read 57344/57344 bytes at offset 70254592 > > +56 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +], > > +"004":[ > > +("write 127M 1M", """wrote 1048576/1048576 bytes at offset 133169152 > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("write 127M 4M", """write failed: Input/output error"""), > > +("write 128M 4096", """write failed: Input/output error"""), > > +("write 140M 4096", """write failed: Input/output error"""), > > +("write -p 140M 4096", """write failed: Input/output error"""), > > +("writev 140M 4096","""writev failed: Input/output error"""), > > +("read 127M 1M", """read 1048576/1048576 bytes at offset 133169152 > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("read 127M 4M", """read failed: Input/output error"""), > > +("read 128M 4096", """read failed: Input/output error"""), > > +("read 140M 4096", """read failed: Input/output error"""), > > +("read -p 140M 4096", """read failed: Input/output error"""), > > +("readv 140M 4096", """readv failed: Input/output error"""), > > +], > > +"008":[ > > +("aio_read 0 128M", """read 134217728/134217728 bytes at offset 0 > > +128 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("aio_write -P 0xa 0 128M", """wrote 134217728/134217728 bytes at offset 0 > > +128 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("aio_read -P 0xa 0 128M", """read 134217728/134217728 bytes at offset 0 > > +128 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +], > > +"025":[ > > +("length", """128 MiB"""), > > +("truncate 384M", """"""), > > +("length", """384 MiB"""), > > +], > > +} > > + > > +io_cases_g = { > > +"016":[ > > +("read -P 0 128M 512", """read 512/512 bytes at offset 134217728 > > +512.000000 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("read -P 0 256M 512", """read 512/512 bytes at offset 268435456 > > +512.000000 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("write -P 66 128M 512", """wrote 512/512 bytes at offset 134217728 > > +512.000000 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("read -P 66 128M 512", """read 512/512 bytes at offset 134217728 > > +512.000000 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("write -P 66 256M 512", """wrote 512/512 bytes at offset 268435456 > > +512.000000 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("read -P 66 256M 512", """read 512/512 bytes at offset 268435456 > > +512.000000 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +], > > +} > > + > > +aio_cases = { > > +"011":[ > > +("""aio_write 1M 1M > > +aio_write 1536K 1M""", """wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) > > +wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("""aio_write 2M 1M > > +aio_write 2560K 1M""", """wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) > > +wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("""aio_write 3M 1M > > +aio_write 3584K 1M""", """wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) > > +wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("""aio_write 4M 1M > > +aio_write 4608K 1M""", """wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) > > +wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("""aio_write 5M 1M > > +aio_write 5632K 1M""", """wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) > > +wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("""aio_write 6M 1M > > +aio_write 6656K 1M""", """wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) > > +wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("""aio_write 7M 1M > > +aio_write 7680K 1M""", """wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) > > +wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("""aio_write 8M 1M > > +aio_write 8704K 1M""", """wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) > > +wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("""aio_write 9M 1M > > +aio_write 9728K 1M""", """wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) > > +wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +("""aio_write 10M 1M > > +aio_write 10752K 1M""", """wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) > > +wrote 1048576/1048576 bytes at offset XXX > > +1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)"""), > > +], > > +} > > + > > +size_cases = { > > +"001":128*1024*1024, > > +"002":128*1024*1024, > > +"003":128*1024*1024, > > +"004":128*1024*1024, > > +"008":128*1024*1024, > > +"011":6*1024*1024*1024, > > +"016":128*1024*1024, > > +"025":128*1024*1024, > > +} > > diff --git a/tests/test_qemu_io.py b/tests/test_qemu_io.py > > new file mode 100644 > > index 0000000..e460f4d > > --- /dev/null > > +++ b/tests/test_qemu_io.py > > @@ -0,0 +1,84 @@ > > +from sheepdog_test import * > > +from testcases import * > > Should be "from qemu_io_testcases import *"? > Oh, my mistake. I'll fix it next version. > > +import time > > + > > +def test_io(): > > + > > + cmd = ["qemu-io"] > > + sdog = Sheepdog() > > + > > + for n in sdog.nodes: > > + n.start() > > + n.wait() > > + > > + p = sdog.format() > > + p.wait() > > + > > + for i in io_cases: > > How about creating a test method for each test case? This test takes > 3 minutes on my environment. > What I was considering is creating a general method for all test cases of one category. So I prefer to split the testcases in the qemu_io_testcases.py. This makes people easier to customize testcases. > > + vdi = sdog.create_vdi(str(i), size_cases[i]) > > + vdi.wait() > > + time.sleep(1) > > + vm = n.create_vm(vdi) > > + for j in io_cases[i]: > > + (out, err) = vm.test_io(cmd, j[0] + "\n") > > + assert out == j[1] > > + time.sleep(1) > > + print "Pass" > > + vdi.destroy() > > + vdi.wait() > > + > > + p = sdog.format() > > + p.wait() > > + for n in sdog.nodes: > > + n.stop() > > + > > +def test_aio(): > > + > > + cmd = ["qemu-io"] > > + sdog = Sheepdog() > > + > > + for n in sdog.nodes: > > + n.start() > > + n.wait() > > + > > + p = sdog.format() > > + p.wait() > > + > > + for i in aio_cases: > > + vdi = sdog.create_vdi(str(i), size_cases[i]) > > + vdi.wait() > > + time.sleep(1) > > + vm = n.create_vm(vdi) > > + for j in aio_cases[i]: > > + (out, err) = vm.test_io(cmd, j[0] + "\n", async=True) > > + assert out == j[1] > > + time.sleep(1) > > + print "Pass" > > + vdi.destroy() > > + vdi.wait() > > + > > +def test_growable_rw(): > > + > > + cmd = ["qemu-io", "-g"] > > + sdog = Sheepdog() > > + > > + for n in sdog.nodes: > > + n.start() > > + n.wait() > > + > > + p = sdog.format() > > + p.wait() > > + > > + for i in io_cases_g: > > + vdi = sdog.create_vdi(str(i), size_cases[i]) > > + vdi.wait() > > + time.sleep(1) > > + vm = n.create_vm(vdi) > > + for j in io_cases_g[i]: > > + (out, err) = vm.test_io(cmd, j[0] + "\n") > > + assert out == j[1] > > + time.sleep(1) > > + print "Pass" > > + vdi.destroy() > > + vdi.wait() > > + > > I'd like to see doing I/O tests under various circumstances (e.g. node > membership is rapidly changing). Do you have a plan to support it? > Not yet. I thought the cases in unit test should be simple and basic. If possible, I think it is better to do such complex test cases (e.g. node membership is rapidly changing) in a single "sheepdog test" project rather than autotest. I'm not an expert on it, but it seems to be better to do so. Thanks for you advice. Baozi > Thanks, > > Kazutaka |