Re: [Qemu-block] [PATCH] iotests: add 248: test resume mirror after auto pause on ENOSPC

[Vladimir Sementsov-Ogievskiy]

20.03.2019 14:03, Kevin Wolf wrote:
> Am 20.03.2019 um 10:59 hat Vladimir Sementsov-Ogievskiy geschrieben:
>> Test that mirror job actually resume on resume command after being
>> automatically paused on ENOSPC error.
>>
>> It's a follow-up test for 8d9648cbf3e
>>      "blockjob: fix user pause in block_job_error_action"
>>
>> Signed-off-by: Vladimir Sementsov-Ogievskiy <address@hidden>
>> ---
>>   tests/qemu-iotests/248     | 77 ++++++++++++++++++++++++++++++++++++++
>>   tests/qemu-iotests/248.out | 12 ++++++
>>   tests/qemu-iotests/group   |  1 +
>>   3 files changed, 90 insertions(+)
>>   create mode 100755 tests/qemu-iotests/248
>>   create mode 100644 tests/qemu-iotests/248.out
>>
>> diff --git a/tests/qemu-iotests/248 b/tests/qemu-iotests/248
>> new file mode 100755
>> index 0000000000..ab88c96532
>> --- /dev/null
>> +++ b/tests/qemu-iotests/248
>> @@ -0,0 +1,77 @@
>> +#!/usr/bin/env python
>> +#
>> +# Test resume mirror after auto pause on ENOSPC
>> +#
>> +# Copyright (c) 2019 Virtuozzo International GmbH. All rights reserved.
>> +#
>> +# 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/>.
>> +#
>> +
>> +import iotests
>> +from iotests import qemu_img_create, qemu_io, file_path, 
>> filter_qmp_testfiles
>> +
>> +iotests.verify_image_format(supported_fmts=['qcow2'])
>> +
>> +source, target = file_path('source', 'target')
>> +size = 5 * 1024 * 1024
>> +limit = 2 * 1024 * 1024
> 
> limit is unused...
> 
>> +qemu_img_create('-f', iotests.imgfmt, source, str(size))
>> +qemu_img_create('-f', iotests.imgfmt, target, str(size))
>> +qemu_io('-c', 'write 0 {}'.format(size), source)
>> +
>> +# raw format don't like empty files
>> +qemu_io('-c', 'write 0 {}'.format(size), target)
>> +
>> +vm = iotests.VM().add_drive(source)
>> +vm.launch()
>> +
>> +vm.qmp_log('blockdev-add', filters=[filter_qmp_testfiles], **{
>> +    'driver': iotests.imgfmt,
>> +    'node-name': 'target',
>> +    'file': {
>> +        'driver': 'raw',
>> +        'size': 2 * 1024 * 1024,
> 
> ...because you used the literal value again here.


hmm, I definitely wanted to use it here)

> 
>> +        'file': {
>> +            'driver': 'file',
>> +            'filename': target
>> +        }
>> +    }
>> +})
>> +
>> +vm.qmp_log('blockdev-mirror', device='drive0', sync='full', target='target',
>> +           on_target_error='enospc')
>> +
>> +vm.event_wait('JOB_STATUS_CHANGE', timeout=3.0,
>> +              match={'data': {'status': 'paused'}})
>> +
>> +vm.qmp_log('query-block-jobs')
>> +
>> +vm.qmp_log('x-blockdev-reopen', filters=[filter_qmp_testfiles], **{
>> +    'driver': iotests.imgfmt,
>> +    'node-name': 'target',
>> +    'file': {
>> +        'driver': 'raw',
>> +        # size parameter dropped
>> +        'file': {
>> +            'driver': 'file',
>> +            'filename': target
>> +        }
>> +    }
>> +})
> 
> Instead of duplicating the configuration here, maybe it would be nicer
> to store it in a variable and just del options['file']['size'] before
> passing them to x-blockdev-reopen.

I thought about this, but it seemed that duplicating in such a small test
looks more visually. On the other hand it leads to questions like yours)
So, OK, will do.

> 
> By the way, I like that you used x-blockdev-reopen. It's much better
> than my suggestion to use HMP qemu-io.
> 
>> +
>> +vm.qmp_log('block-job-resume', device='drive0')
>> +vm.qmp_log('query-block-jobs')  # check that job is resumed
> 
> This is the expected output for this query-block-jobs:
> 
>> +{"return": [{"auto-dismiss": true, "auto-finalize": true, "busy": true, 
>> "device": "drive0", "io-status": "ok", "len": 5242880, "offset": 1048576, 
>> "paused": false, "ready": false, "speed": 0, "status": "running", "type": 
>> "mirror"}]}
> 
> So it expects a specific offset of 1048576. Is this reliable without
> throttling the job?

Hmm, don't know. May be better to filter offset out.

> 
> Kevin
> 


-- 
Best regards,
Vladimir