[Sheepdog] [PATCH 2/2] tests: add qemu-io testcases.
Chen Baozi
chenbaozi at gmail.com
Mon Nov 14 04:37:45 CET 2011
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
More information about the sheepdog
mailing list