xemu/qapi
Kevin Wolf ffa244c84a file-posix: Mitigate file fragmentation with extent size hints
Especially when O_DIRECT is used with image files so that the page cache
indirection can't cause a merge of allocating requests, the file will
fragment on the file system layer, with a potentially very small
fragment size (this depends on the requests the guest sent).

On Linux, fragmentation can be reduced by setting an extent size hint
when creating the file (at least on XFS, it can't be set any more after
the first extent has been allocated), basically giving raw files a
"cluster size" for allocation.

This adds a create option to set the extent size hint, and changes the
default from not setting a hint to setting it to 1 MB. The main reason
why qcow2 defaults to smaller cluster sizes is that COW becomes more
expensive, which is not an issue with raw files, so we can choose a
larger size. The tradeoff here is only potentially wasted disk space.

For qcow2 (or other image formats) over file-posix, the advantage should
even be greater because they grow sequentially without leaving holes, so
there won't be wasted space. Setting even larger extent size hints for
such images may make sense. This can be done with the new option, but
let's keep the default conservative for now.

The effect is very visible with a test that intentionally creates a
badly fragmented file with qemu-img bench (the time difference while
creating the file is already remarkable) and then looks at the number of
extents and the time a simple "qemu-img map" takes.

Without an extent size hint:

    $ ./qemu-img create -f raw -o extent_size_hint=0 ~/tmp/test.raw 10G
    Formatting '/home/kwolf/tmp/test.raw', fmt=raw size=10737418240 extent_size_hint=0
    $ ./qemu-img bench -f raw -t none -n -w ~/tmp/test.raw -c 1000000 -S 8192 -o 0
    Sending 1000000 write requests, 4096 bytes each, 64 in parallel (starting at offset 0, step size 8192)
    Run completed in 25.848 seconds.
    $ ./qemu-img bench -f raw -t none -n -w ~/tmp/test.raw -c 1000000 -S 8192 -o 4096
    Sending 1000000 write requests, 4096 bytes each, 64 in parallel (starting at offset 4096, step size 8192)
    Run completed in 19.616 seconds.
    $ filefrag ~/tmp/test.raw
    /home/kwolf/tmp/test.raw: 2000000 extents found
    $ time ./qemu-img map ~/tmp/test.raw
    Offset          Length          Mapped to       File
    0               0x1e8480000     0               /home/kwolf/tmp/test.raw

    real    0m1,279s
    user    0m0,043s
    sys     0m1,226s

With the new default extent size hint of 1 MB:

    $ ./qemu-img create -f raw -o extent_size_hint=1M ~/tmp/test.raw 10G
    Formatting '/home/kwolf/tmp/test.raw', fmt=raw size=10737418240 extent_size_hint=1048576
    $ ./qemu-img bench -f raw -t none -n -w ~/tmp/test.raw -c 1000000 -S 8192 -o 0
    Sending 1000000 write requests, 4096 bytes each, 64 in parallel (starting at offset 0, step size 8192)
    Run completed in 11.833 seconds.
    $ ./qemu-img bench -f raw -t none -n -w ~/tmp/test.raw -c 1000000 -S 8192 -o 4096
    Sending 1000000 write requests, 4096 bytes each, 64 in parallel (starting at offset 4096, step size 8192)
    Run completed in 10.155 seconds.
    $ filefrag ~/tmp/test.raw
    /home/kwolf/tmp/test.raw: 178 extents found
    $ time ./qemu-img map ~/tmp/test.raw
    Offset          Length          Mapped to       File
    0               0x1e8480000     0               /home/kwolf/tmp/test.raw

    real    0m0,061s
    user    0m0,040s
    sys     0m0,014s

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20200707142329.48303-1-kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2020-07-14 15:18:59 +02:00
..
audio.json audio/jack: add JACK client audiodev 2020-05-25 11:30:03 +02:00
authz.json authz: add QAuthZList object type for an access control list 2019-02-26 15:32:18 +00:00
block-core.json file-posix: Mitigate file fragmentation with extent size hints 2020-07-14 15:18:59 +02:00
block.json qapi: Mark deprecated QMP parts with feature 'deprecated' 2020-03-17 21:43:12 +01:00
char.json qapi: Belatedly update doc comment for @wait deprecation 2020-03-17 19:58:34 +01:00
common.json qapi: Move query-target from misc.json to machine.json 2019-08-21 13:24:01 +02:00
control.json qapi: Mark deprecated QMP parts with feature 'deprecated' 2020-03-17 21:43:12 +01:00
crypto.json qcrypto/luks: implement encryption key management 2020-07-06 08:49:28 +02:00
dump.json qapi: Fix indent level on doc comments in json files 2020-02-15 11:41:50 +01:00
error.json qapi: Split error.json off common.json 2019-08-16 13:31:51 +02:00
introspect.json qapi: Add feature flags to struct members 2020-03-17 21:25:47 +01:00
job.json block/core: add generic infrastructure for x-blockdev-amend qmp command 2020-07-06 08:49:28 +02:00
machine-target.json qapi: Fix indent level on doc comments in json files 2020-02-15 11:41:50 +01:00
machine.json target/avr: Register AVR support with the rest of QEMU 2020-07-11 11:02:05 +02:00
Makefile.objs qemu-storage-daemon: Add --monitor option 2020-03-06 17:21:28 +01:00
migration.json migration/xbzrle: add encoding rate 2020-05-07 17:40:24 +01:00
misc-target.json qapi: Fix indent level on doc comments in json files 2020-02-15 11:41:50 +01:00
misc.json virtio-pci: Send qapi events when the virtio-mem size changes 2020-07-03 07:57:04 -04:00
net.json vhost-vdpa: introduce vhost-vdpa net client 2020-07-07 07:59:51 -04:00
opts-visitor.c qapi: Make visitor functions taking Error ** return bool, not void 2020-07-10 15:18:08 +02:00
pragma.json qapi: Create 'pragma' module 2020-03-06 17:21:28 +01:00
qapi-clone-visitor.c qapi: Make visitor functions taking Error ** return bool, not void 2020-07-10 15:18:08 +02:00
qapi-dealloc-visitor.c qapi: Make visitor functions taking Error ** return bool, not void 2020-07-10 15:18:08 +02:00
qapi-schema.json qapi: Create 'pragma' module 2020-03-06 17:21:28 +01:00
qapi-util.c qemu-common: Move qemu_isalnum() etc. to qemu/ctype.h 2019-06-11 20:22:09 +02:00
qapi-visit-core.c qapi: Purge error_propagate() from QAPI core 2020-07-10 15:18:08 +02:00
qdev.json qapi: Fix indent level on doc comments in json files 2020-02-15 11:41:50 +01:00
qmp-dispatch.c qmp: fix leak on callbacks that return both value and error 2020-04-02 14:55:46 -04:00
qmp-event.c Include qemu-common.h exactly where needed 2019-06-12 13:20:20 +02:00
qmp-registry.c qmp: constify QmpCommand and list 2020-03-17 21:43:12 +01:00
qobject-input-visitor.c qapi: Make visitor functions taking Error ** return bool, not void 2020-07-10 15:18:08 +02:00
qobject-output-visitor.c qapi: Make visitor functions taking Error ** return bool, not void 2020-07-10 15:18:08 +02:00
qom.json qapi: Flatten object-add 2020-03-06 17:21:27 +01:00
rdma.json qapi: Define new QMP message for pvrdma 2018-12-22 11:09:56 +02:00
rocker.json qapi: Fix indent level on doc comments in json files 2020-02-15 11:41:50 +01:00
run-state.json qapi: Fix indent level on doc comments in json files 2020-02-15 11:41:50 +01:00
sockets.json qemu-sockets: add abstract UNIX domain socket support 2020-05-20 10:34:40 +01:00
string-input-visitor.c error: Eliminate error_propagate() with Coccinelle, part 1 2020-07-10 15:18:08 +02:00
string-output-visitor.c qapi: Make visitor functions taking Error ** return bool, not void 2020-07-10 15:18:08 +02:00
tpm.json qapi: Use explicit bulleted lists 2020-02-15 11:41:50 +01:00
trace-events trace-events: Shorten file names in comments 2019-03-22 16:18:07 +00:00
trace.json qapi: Add blank lines before bulleted lists 2020-02-15 11:41:50 +01:00
transaction.json qemu-storage-daemon: Add --monitor option 2020-03-06 17:21:28 +01:00
ui.json qapi: Delete all the "foo: dropped in n.n" notes 2020-02-15 11:41:50 +01:00