Re: [sheepdog] [PATCH] sheepdog proto: avoid using stdint.h macros, so sbd can build.

On 2015/1/20 10:19, Hitoshi Mitake wrote: At Mon, 19 Jan 2015 12:39:44 -0800, Alexander Guy wrote: The Linux kernel headers don't have the *_C suffix macros, so they can't be used in a header file shared between sheepdog userspace and the sbd kernel module. Signed-off-by: Alexander Guy

Re: [sheepdog] Issues about setup

At Tue, 20 Jan 2015 10:46:38 +0800, hujianyang wrote: On 2015/1/20 10:32, Hitoshi Mitake wrote: I did a little change and it compiles OK. But UINT*_C are widely used in sheepdog_proto.h so maybe a large-scale modification is needed? Cc-ing Alexander. The above UINT*_C macro problem

[sheepdog] kill one node from cluster will get a lot of errors

6 nodes sheepdog cluster (v0.9.1) When I kill one node, I get the following error message. When I restart the close node, I get the same error message. Jan 20 11:17:33 INFO [main] recover_object_main(905) object recovery progress 97% Jan 20 11:17:33 ALERT [rw 22948] get_vdi_copy_number(110)

Re: [sheepdog] [PATCH] sheepdog proto: avoid using stdint.h macros, so sbd can build.

At Mon, 19 Jan 2015 12:39:44 -0800, Alexander Guy wrote: The Linux kernel headers don't have the *_C suffix macros, so they can't be used in a header file shared between sheepdog userspace and the sbd kernel module. Signed-off-by: Alexander Guy alexan...@andern.org ---

Re: [sheepdog] [PATCH v2 0/4] deep copied snapshot

At Mon, 19 Jan 2015 19:28:50 +0900, Hitoshi Mitake wrote: This patchset implements a new feature: deep copied snapshot. In a case of long living, periodically snapshotted VDIs, it is difficult to collect garbage inodes because the case can produce a large family of VDIs. This patch adds a

Re: [sheepdog] Issues about setup

At Mon, 19 Jan 2015 16:29:17 +0800, hujianyang wrote: Hi Hitoshi, On 2015/1/16 17:40, Hitoshi Mitake wrote: Hmm, could you show your yasm version? My yasm (1.2.0) can build current master branch with no problems. Thanks, Hitoshi Thanks for your advise. I've updated my

Re: [sheepdog] Issues about setup

On 2015/1/20 10:32, Hitoshi Mitake wrote: I did a little change and it compiles OK. But UINT*_C are widely used in sheepdog_proto.h so maybe a large-scale modification is needed? Cc-ing Alexander. The above UINT*_C macro problem is solved by the recent patch of Alexander. But the kstrtoul()

Re: [sheepdog] [PATCH] sheepdog proto: avoid using stdint.h macros, so sbd can build.

At Tue, 20 Jan 2015 10:26:33 +0800, hujianyang wrote: On 2015/1/20 10:19, Hitoshi Mitake wrote: At Mon, 19 Jan 2015 12:39:44 -0800, Alexander Guy wrote: The Linux kernel headers don't have the *_C suffix macros, so they can't be used in a header file shared between sheepdog userspace

[sheepdog] [PATCH] sbd: use kstrtoul() instead of strict_strtoul()

Cc: hujianyang hujiany...@huawei.com Signed-off-by: Hitoshi Mitake mitake.hito...@lab.ntt.co.jp --- sbd/sheep_block_device.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sbd/sheep_block_device.c b/sbd/sheep_block_device.c index eaee932..d82ff43 100644 ---

[sheepdog] Build failed in Jenkins: sheepdog-build #608

See http://jenkins.sheepdog-project.org:8080/job/sheepdog-build/608/ -- [...truncated 57 lines...] checking for grep that handles long lines and -e... /bin/grep checking for egrep... /bin/grep -E checking for ANSI C header files... yes checking for

Re: [sheepdog] [PATCH] sbd: use kstrtoul() instead of strict_strtoul()

At Tue, 20 Jan 2015 11:46:32 +0900, Hitoshi Mitake wrote: Cc: hujianyang hujiany...@huawei.com Signed-off-by: Hitoshi Mitake mitake.hito...@lab.ntt.co.jp --- sbd/sheep_block_device.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) Dropping it. The version information should be

Re: [sheepdog] Issues about setup

On 2015/1/20 10:49, Hitoshi Mitake wrote: At Tue, 20 Jan 2015 10:46:38 +0800, hujianyang wrote: On 2015/1/20 10:32, Hitoshi Mitake wrote: I did a little change and it compiles OK. But UINT*_C are widely used in sheepdog_proto.h so maybe a large-scale modification is needed? Cc-ing

Re: [sheepdog] [PATCH v2 3/4] sheep, dog: fast deep copy for snapshot

At Mon, 19 Jan 2015 16:23:34 +0400, Vasiliy Tolstov wrote: 2015-01-19 13:28 GMT+03:00 Hitoshi Mitake mitake.hito...@lab.ntt.co.jp: dog vdi snapshot vdi --no-share has a bottleneck: the dog process. This patch adds a new option --fast-deep-copy to dog vdi snapshot, which avoid the

[sheepdog] 答复: [PATCH] sbd: use kstrtoul() instead of strict_strtoul()

6 nodes sheepdog cluster (v0.9.1) When I kill one node, I get the following error message. When I restart the close node, I get the same error message. Jan 20 11:17:33 INFO [main] recover_object_main(905) object recovery progress 97% Jan 20 11:17:33 ALERT [rw 22948] get_vdi_copy_number(110)

Re: [sheepdog] [PATCH] sbd: use kstrtoul() instead of strict_strtoul()

On 2015/1/20 11:01, Hitoshi Mitake wrote: At Tue, 20 Jan 2015 11:46:32 +0900, Hitoshi Mitake wrote: Cc: hujianyang hujiany...@huawei.com Signed-off-by: Hitoshi Mitake mitake.hito...@lab.ntt.co.jp --- sbd/sheep_block_device.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)

Re: [sheepdog] Issues about setup

At Tue, 20 Jan 2015 11:00:53 +0800, hujianyang wrote: On 2015/1/20 10:49, Hitoshi Mitake wrote: At Tue, 20 Jan 2015 10:46:38 +0800, hujianyang wrote: On 2015/1/20 10:32, Hitoshi Mitake wrote: I did a little change and it compiles OK. But UINT*_C are widely used in sheepdog_proto.h

[sheepdog] [PATCH v2] sheepdog: selectable object size support

Previously, qemu block driver of sheepdog used hard-coded VDI object size. This patch enables users to handle block_size_shift value for calculating VDI object size. When you start qemu, you don't need to specify additional command option. But when you create the VDI which doesn't have default

[sheepdog] [PATCH v2 3/4] sheep, dog: fast deep copy for snapshot

dog vdi snapshot vdi --no-share has a bottleneck: the dog process. This patch adds a new option --fast-deep-copy to dog vdi snapshot, which avoid the bottleneck. If the option is passed to the dog command, actual copying of data objects are done by sheep processes. So the dog process isn't a

[sheepdog] [PATCH v2 4/4] tests/functional: add a new test for checking fast deep copied snapshot

Signed-off-by: Hitoshi Mitake mitake.hito...@lab.ntt.co.jp --- tests/functional/101 | 26 ++ tests/functional/101.out | 16 tests/functional/group | 1 + 3 files changed, 43 insertions(+) create mode 100755 tests/functional/101 create mode 100644

Re: [sheepdog] [PATCH v2 3/4] sheep, dog: fast deep copy for snapshot

2015-01-19 13:28 GMT+03:00 Hitoshi Mitake mitake.hito...@lab.ntt.co.jp: dog vdi snapshot vdi --no-share has a bottleneck: the dog process. This patch adds a new option --fast-deep-copy to dog vdi snapshot, which avoid the bottleneck. If the option is passed to the dog command, actual copying

Re: [sheepdog] Issues about setup

Hi Hitoshi, On 2015/1/16 17:40, Hitoshi Mitake wrote: Hmm, could you show your yasm version? My yasm (1.2.0) can build current master branch with no problems. Thanks, Hitoshi Thanks for your advise. I've updated my yasm to 1.3.0 and found it's OK. I think maybe we should mark this

[sheepdog] [PATCH v2 1/4] sheep, dog: support deepcopy snapshot

Current sheepdog supports deepcopy cloning of VDI (with --no-share option). This patch also implements the deepcopy feature for snapshot. It is useful for fine grained GC of VID which is important for long living, periodically snapshotted VDIs. This patch introduces two important changes: 1.

[sheepdog] [PATCH v2 0/4] deep copied snapshot

This patchset implements a new feature: deep copied snapshot. In a case of long living, periodically snapshotted VDIs, it is difficult to collect garbage inodes because the case can produce a large family of VDIs. This patch adds a new option --no-share (like vdi cloning) to dog vdi snapshot.

[sheepdog] [PATCH v2 2/4] tests/functional: add a test for deep copied snapshots

Signed-off-by: Hitoshi Mitake mitake.hito...@lab.ntt.co.jp --- tests/functional/100 | 50 tests/functional/100.out | 39 + tests/functional/group | 1 + 3 files changed, 90 insertions(+) create mode

[sheepdog] [PATCH] sheepdog proto: avoid using stdint.h macros, so sbd can build.

The Linux kernel headers don't have the *_C suffix macros, so they can't be used in a header file shared between sheepdog userspace and the sbd kernel module. Signed-off-by: Alexander Guy alexan...@andern.org --- include/sheepdog_proto.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)