On Sun, Aug 02, 2015 at 02:52:08PM +0300, Vasiliy Tolstov wrote:
2015-07-31 15:08 GMT+03:00 Vasiliy Tolstov v.tols...@selfip.ru:
Please wait to performance comparison. As i see Liu's patch may be
more slow then Hitoshi.
I'm switch to local cluster driver to test only local ssd and not
On Fri, Jul 31, 2015 at 03:08:09PM +0300, Vasiliy Tolstov wrote:
2015-07-31 14:55 GMT+03:00 Vasiliy Tolstov v.tols...@selfip.ru:
Liu's patch also works for me. But also like in Hitoshi patch breaks
when using discards in qemu =(.
Please wait to performance comparison. As i see Liu's
On Wed, Jul 29, 2015 at 02:04:55PM +0900, Hitoshi Mitake wrote:
At Wed, 29 Jul 2015 12:02:35 +0800,
Liu Yuan wrote:
From: Liu Yuan liuy...@cmss.chinamobile.com
Current sheepdog driver use a range update_inode(min_idx, max_idx) for
batching
the updates. But there is subtle problem
On Wed, Jul 29, 2015 at 10:17:45AM +0800, 陈李粮 wrote:
what is sheepdog-ng?
It is a forked of Sheepdog due to unstability and unfriendliness of open source
project. Sheepdog, in my eyes, now become a NTT's project, so Kazutaka and I
quit the maintainership of sheepdog and moved to sheepdog-ng.
On Sat, Jul 18, 2015 at 01:44:24AM +0900, Hitoshi Mitake wrote:
Current sheepdog driver only serializes create requests in oid
unit. This mechanism isn't enough for handling requests to
overwrapping area spanning multiple oids, so it can result bugs like
below:
On Tue, Jul 28, 2015 at 04:50:08PM +0800, Liu Yuan wrote:
On Sat, Jul 18, 2015 at 01:44:24AM +0900, Hitoshi Mitake wrote:
Current sheepdog driver only serializes create requests in oid
unit. This mechanism isn't enough for handling requests to
overwrapping area spanning multiple oids, so
On Mon, Jul 27, 2015 at 11:23:02AM -0400, Jeff Cody wrote:
On Sat, Jul 18, 2015 at 01:44:24AM +0900, Hitoshi Mitake wrote:
Current sheepdog driver only serializes create requests in oid
unit. This mechanism isn't enough for handling requests to
overwrapping area spanning multiple oids, so
On Tue, Jul 28, 2015 at 10:31:32PM +0800, Liu Yuan wrote:
On Mon, Jul 27, 2015 at 11:23:02AM -0400, Jeff Cody wrote:
On Sat, Jul 18, 2015 at 01:44:24AM +0900, Hitoshi Mitake wrote:
Current sheepdog driver only serializes create requests in oid
unit. This mechanism isn't enough
From: Liu Yuan liuy...@cmss.chinamobile.com
Current sheepdog driver use a range update_inode(min_idx, max_idx) for batching
the updates. But there is subtle problem by determining min_idx and max_idx:
for a single create request, min_idx == max_idx, so actually we just update one
one bit
On Wed, Jul 08, 2015 at 11:10:47AM +0800, Yu Yang wrote:
There is a struct node_id which saves the sheep node address and port as
follows
in internal_proto.h:
struct node_id {
uint8_t addr[16];
uint16_t port;
uint8_t io_addr[16];
uint16_t io_port;
uint8_t
On Wed, May 27, 2015 at 04:24:29PM +0800, yuyang wrote:
Hello everyone,
Recently we need copy one VDI from sheepdog to another storage
system(such as sheepdog or RAID). There are 2 ways to choose,
1. dog vdi r | dog vdi w,(only for sheepdog, or temp file is needed)
2. lib client, which is
On Thu, May 14, 2015 at 03:12:35PM +0800, Yu Yang wrote:
From: Yu Yang yuy...@cmss.chinamobile.com
The return value should be SD_RES_XXX, so return SD_RES_SYSTEM_ERROR
when sheep_submit_sdreq() failed.
Signed-off-by: Yu Yang yuy...@cmss.chinamobile.com
---
lib/shared/sheep.c |2 +-
On Thu, May 14, 2015 at 10:31:47PM +0300, Vasiliy Tolstov wrote:
Per volume redundancy useful for hostings or openstack based solutions. So
users may need to pay for stability or increase speed..
14 мая 2015 г. 13:21 пользователь Valerio Pachera siri...@gmail.com
написал:
IMHO, per volume
On Mon, May 18, 2015 at 09:52:03AM +0900, Hitoshi Mitake wrote:
At Thu, 14 May 2015 00:57:24 +0800,
Liu Yuan wrote:
Hi y'all,
Based on recent frustrating[1] debug on our test cluster, I'd like propsose,
which might looks very radical to you, that we should remove struct
vdi_state
On Mon, May 18, 2015 at 01:42:59PM +0900, Hitoshi Mitake wrote:
At Mon, 18 May 2015 09:46:25 +0800,
Liu Yuan wrote:
On Mon, May 18, 2015 at 09:52:03AM +0900, Hitoshi Mitake wrote:
At Thu, 14 May 2015 00:57:24 +0800,
Liu Yuan wrote:
Hi y'all,
Based on recent
Hi y'all,
Based on recent frustrating[1] debug on our test cluster, I'd like propsose,
which might looks very radical to you, that we should remove struct vdi_state
completely from sheepdog code.
Let me show you the background picture how it was introduced. It was
introduced by Leven Li by the
From: Liu Yuan liuy...@cmss.chinamobile.com
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
sbd/sheep.c | 13 -
1 file changed, 4 insertions(+), 9 deletions(-)
diff --git a/sbd/sheep.c b/sbd/sheep.c
index bfc935b..8b74b59 100644
--- a/sbd/sheep.c
+++ b/sbd/sheep.c
@@ -34,7
From: Liu Yuan liuy...@cmss.chinamobile.com
This is a preparation patch for creating user statically linked libsheepdog.a.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
dog/Makefile.am | 4 ++--
include/common.h | 2 +-
lib/Makefile.am | 14
From: Liu Yuan liuy...@cmss.chinamobile.com
Some use case might want to statically linked to the sheepdog library.
Howto:
$ gcc your_prog.c -lpthread /path/to/libsheepdog.a
Note, 'make install' will install libsheepdog.a into /usr/lib/ along with
libsheepdog.so.
Signed-off-by: Liu Yuan liuy
On Mon, Apr 27, 2015 at 12:54:22PM +0900, Hitoshi Mitake wrote:
From: Masaki Saeki saeki.mas...@po.ntts.co.jp
This change is a preparation patch for add store_driver.
Put files together to the new folder.
Signed-off-by: Masaki Saeki saeki.mas...@po.ntts.co.jp
Applied this series, thanks.
From: Liu Yuan liuy...@cmss.chinamobile.com
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
lib/shared/sheepdog.h | 70 +--
1 file changed, 68 insertions(+), 2 deletions(-)
diff --git a/lib/shared/sheepdog.h b/lib/shared/sheepdog.h
index
From: Liu Yuan liuy...@cmss.chinamobile.com
internal.h is dedicated to hold the non-user functions and structures for shared
lib.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
lib/shared/internal.h | 28
lib/shared/sheep.c| 1 +
lib/shared/sheepdog.h
From: Liu Yuan liuy...@cmss.chinamobile.com
Have all the sd_xxx functions return SD_RES_XXX, we will have a unified error
handling for user program and more importantly, std error can't cover all the
error cases that is specific to the sheepdog cluster, such as SD_RES_NO_VDI
and SD_RES_NO_OBJ
From: Liu Yuan liuy...@cmss.chinamobile.com
internal.h is dedicated to hold the non-user functions and structures for shared
lib.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
lib/shared/internal.h | 28
lib/shared/sheep.c| 1 +
lib/shared/sheepdog.h
From: Liu Yuan liuy...@cmss.chinamobile.com
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
lib/shared/sheepdog.h | 70 +--
1 file changed, 68 insertions(+), 2 deletions(-)
diff --git a/lib/shared/sheepdog.h b/lib/shared/sheepdog.h
index
On Mon, Apr 13, 2015 at 01:32:56PM +0300, Vasiliy Tolstov wrote:
sheepdog now able to run in foreground, so use this in systemd service,
also remove running shell script to get defaults, because it provided
in service file now.
Signed-off-by: Vasiliy Tolstov v.tols...@selfip.ru
---
On Mon, Apr 13, 2015 at 02:35:02PM +0800, Ruoyu wrote:
This patch fixes the compiler warning message as below.
warning: implicit declaration of function ‘install_sighandler’
[-Wimplicit-function-declaration]
Signed-off-by: Ruoyu lian...@ucweb.com
---
lib/work.c | 1 +
1 file changed, 1
On Sun, Apr 12, 2015 at 07:38:09PM +0800, passedwind wrote:
today i change my sheepdog cluster data area node network with bonding mode
4 with (Bonding Mode: IEEE 802.3ad Dynamic link aggregation).gateway node
with normal network interface. gateway node to dataArea node network link no
On Thu, Apr 09, 2015 at 03:30:21PM +0800, 张灿群 wrote:
dear,
When i start sheep daemon with journal,the daemon couldn't start.
command:
sheep -j dir=/sdac,size=160G,skip=1 -l dir=/var/log/,level=err,format=server
/sdad/ -n -D -y 10.64.0.38 -p 7000 -c
On Wed, Apr 08, 2015 at 09:14:29AM +0300, Vasiliy Tolstov wrote:
Signed-off-by: Vasiliy Tolstov v.tols...@selfip.ru
---
sheep/sheep.c | 26 +++--
1 file changed, 14 insertions(+), 12 deletions(-)
diff --git a/sheep/sheep.c b/sheep/sheep.c
index 86444b0..38ffed0
On Wed, Apr 08, 2015 at 09:26:05AM +0300, Vasiliy Tolstov wrote:
sheepdog now able to run in foreground, so use this in systemd service,
also remove running shell script to get defaults, because it provided
in service file now.
Signed-off-by: Vasiliy Tolstov v.tols...@selfip.ru
---
On Wed, Apr 08, 2015 at 09:43:10AM +0300, Vasiliy Tolstov wrote:
Signed-off-by: Vasiliy Tolstov v.tols...@selfip.ru
---
sheep/sheep.c | 26 +++--
1 file changed, 14 insertions(+), 12 deletions(-)
diff --git a/sheep/sheep.c b/sheep/sheep.c
index 86444b0..4f6a662
On Wed, Apr 08, 2015 at 01:21:05AM +0900, Hitoshi Mitake wrote:
At Fri, 3 Apr 2015 11:20:48 +0800,
Liu Yuan wrote:
From: Liu Yuan liuy...@cmss.chinamobile.com
This is a prepation patch to make util.c self-contained and then can be used
outside sheep.
Signed-off-by: Liu Yuan
On Wed, Apr 08, 2015 at 01:19:20AM +0900, Hitoshi Mitake wrote:
At Fri, 3 Apr 2015 11:20:54 +0800,
Liu Yuan wrote:
From: Liu Yuan liuy...@cmss.chinamobile.com
Add the final makefile stuff
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
configure.ac| 3
From: Liu Yuan liuy...@cmss.chinamobile.com
This is a prepation patch to make util.c self-contained and then can be used
outside sheep.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
dog/common.c| 25 +
dog/dog.h | 5 +
include/Makefile.am | 3 +-
include
From: Liu Yuan liuy...@cmss.chinamobile.com
This is a preparation patch to dissociate util.h and utli.c from sheepdog,
in order to be used by shared lib.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
configure.ac | 11 ---
include/logger.h | 7
From: Liu Yuan liuy...@cmss.chinamobile.com
The core idea is the same as sbd's framework.
User request
|
| aio control block (struct aiocb)
V
+++
|||
r1 r2r3
|||
VVV
From: Liu Yuan liuy...@cmss.chinamobile.com
Finally, we kick started, though it is far away from a complete implemention.
This patch set introduces a shared library and for now it only supports simple
vdi operation. The vdi read/write operation would be the most difficult part of
shared library
From: Liu Yuan liuy...@cmss.chinamobile.com
This is preparation patch for shared lib.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
include/sheep.h | 13 +
include/sheepdog_proto.h | 16 +---
lib/sd_inode.c | 1 +
3 files changed, 15
From: Liu Yuan liuy...@cmss.chinamobile.com
This is a preparation patch. list.h and util.h will be used by shared library
and exported to the use.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
include/compiler.h | 9 -
include/list.h | 2 --
include/util.h | 14
From: Liu Yuan liuy...@cmss.chinamobile.com
This is a preparation patch for shared lib. shared lib can't call sd logger.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
dog/dog.h| 1 +
dog/treeview.c | 1 +
include/fec.h| 1 +
include/logger.h | 6 --
include
On Fri, Apr 03, 2015 at 03:59:18PM +0900, Hitoshi Mitake wrote:
At Fri, 3 Apr 2015 12:06:56 +0800,
Liu Yuan wrote:
On Fri, Apr 03, 2015 at 12:49:31PM +0900, Hitoshi Mitake wrote:
At Fri, 3 Apr 2015 11:20:46 +0800,
Liu Yuan wrote:
From: Liu Yuan liuy...@cmss.chinamobile.com
From: Liu Yuan liuy...@cmss.chinamobile.com
This will kill a hidden build problem because of stale .o in libsheepdog.a
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
lib/Makefile.am | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/lib/Makefile.am b/lib/Makefile.am
From: Liu Yuan liuy...@cmss.chinamobile.com
Finally, we kick started, though it is far away from a complete implemention.
This patch set introduces a shared library and for now it only supports simple
vdi operation. The vdi read/write operation would be the most difficult part of
shared library
From: Liu Yuan liuy...@cmss.chinamobile.com
The core idea is the same as sbd's framework.
User request
|
| aio control block (struct aiocb)
V
+++
|||
r1 r2r3
|||
VVV
From: Liu Yuan liuy...@cmss.chinamobile.com
This is a prepation patch to make util.c self-contained and then can be used
outside sheep.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
dog/common.c| 25 +
dog/dog.h | 5 +
include/Makefile.am | 3 +-
include
From: Liu Yuan liuy...@cmss.chinamobile.com
This is a preparation patch to dissociate util.h and utli.c from sheepdog,
in order to be used by shared lib.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
configure.ac | 11 ---
include/logger.h | 7
From: Liu Yuan liuy...@cmss.chinamobile.com
This is preparation patch for shared lib.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
include/sheep.h | 13 +
include/sheepdog_proto.h | 16 +---
lib/sd_inode.c | 1 +
3 files changed, 15
From: Liu Yuan liuy...@cmss.chinamobile.com
This is a preparation patch for shared lib. shared lib can't call sd logger.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
dog/dog.h| 1 +
dog/treeview.c | 1 +
include/fec.h| 1 +
include/logger.h | 6 --
include
From: Liu Yuan liuy...@cmss.chinamobile.com
This is a preparation patch. list.h and util.h will be used by shared library
and exported to the use.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
include/compiler.h | 9 -
include/list.h | 2 --
include/util.h | 14
From: Liu Yuan liuy...@cmss.chinamobile.com
Add the final makefile stuff
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
configure.ac| 3 ++-
lib/Makefile.am | 14 ++
2 files changed, 16 insertions(+), 1 deletion(-)
diff --git a/configure.ac b/configure.ac
index
From: Liu Yuan liuy...@cmss.chinamobile.com
Nothing intersting, these functions are basically wrappers of sheep.c.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
lib/shared/sheep.c| 3 +-
lib/shared/sheepdog.h | 10 ++-
lib/shared/vdi.c | 185
On Fri, Apr 03, 2015 at 12:49:31PM +0900, Hitoshi Mitake wrote:
At Fri, 3 Apr 2015 11:20:46 +0800,
Liu Yuan wrote:
From: Liu Yuan liuy...@cmss.chinamobile.com
Finally, we kick started, though it is far away from a complete
implemention.
This patch set introduces a shared
On Fri, Apr 03, 2015 at 12:06:56PM +0800, Liu Yuan wrote:
On Fri, Apr 03, 2015 at 12:49:31PM +0900, Hitoshi Mitake wrote:
At Fri, 3 Apr 2015 11:20:46 +0800,
Liu Yuan wrote:
From: Liu Yuan liuy...@cmss.chinamobile.com
Finally, we kick started, though it is far away from
On Wed, Mar 25, 2015 at 11:33:17AM +0300, Vasiliy Tolstov wrote:
2015-03-25 5:13 GMT+03:00 Liu Yuan namei.u...@gmail.com:
If you just want to see log in stdout, you can try -l dst=stdout'. Any other
reason to bring this back?
firstly under systemd forking not recommended, also you
On Tue, Mar 17, 2015 at 05:30:47PM +0300, Vasiliy Tolstov wrote:
Signed-off-by: Vasiliy Tolstov v.tols...@selfip.ru
---
sheep/sheep.c | 29 +
1 file changed, 17 insertions(+), 12 deletions(-)
diff --git a/sheep/sheep.c b/sheep/sheep.c
index fc53ec9..cad8293
with -l dst=stdout.
Signed-off-by: Hitoshi Mitake mitake.hito...@lab.ntt.co.jp
Signed-off-by: Liu Yuan namei.u...@gmail.com
---
If you just want to see log in stdout, you can try -l dst=stdout'. Any other
reason to bring
On Tue, Mar 17, 2015 at 05:30:47PM +0300, Vasiliy Tolstov wrote:
Signed-off-by: Vasiliy Tolstov v.tols...@selfip.ru
---
sheep/sheep.c | 29 +
1 file changed, 17 insertions(+), 12 deletions(-)
diff --git a/sheep/sheep.c b/sheep/sheep.c
index fc53ec9..cad8293
On Mon, Mar 23, 2015 at 02:18:24PM +0900, Saeki Masaki wrote:
On 2015/03/21 21:02, Liu Yuan wrote:
On Fri, Mar 20, 2015 at 06:39:08PM +0900, Saeki Masaki wrote:
Current sheepdog stores whole objects in single directory like
/var/lib/sheepdog/obj
This mechanism is difficult to handle massive
On Fri, Mar 20, 2015 at 06:39:08PM +0900, Saeki Masaki wrote:
Current sheepdog stores whole objects in single directory like
/var/lib/sheepdog/obj
This mechanism is difficult to handle massive files when increasing cluster
volume.
In particular, inode object having special informations
sheep_SOURCES+= http/http.c http/kv.c http/s3.c http/swift.c
\
diff --git a/sheep/md.c b/sheep/md.c
deleted file mode 100644
index 87ab759..000
--- a/sheep/md.c
+++ /dev/null
@@ -1,878 +0,0 @@
-/*
- * Copyright (C) 2013 Taobao Inc.
- *
- * Liu Yuan namei.u...@gmail.com
On Tue, Mar 17, 2015 at 02:51:32PM +0900, Hitoshi Mitake wrote:
At Tue, 17 Mar 2015 13:40:51 +0800,
Liu Yuan wrote:
On Tue, Mar 17, 2015 at 01:32:44PM +0800, Liu Yuan wrote:
On Tue, Mar 17, 2015 at 01:16:00PM +0800, Liu Yuan wrote:
On Tue, Mar 17, 2015 at 02:03:03PM +0900, Hitoshi
On Tue, Mar 17, 2015 at 02:59:12PM +0900, Hitoshi Mitake wrote:
At Tue, 17 Mar 2015 13:54:16 +0800,
Liu Yuan wrote:
On Tue, Mar 17, 2015 at 01:40:51PM +0800, Liu Yuan wrote:
On Tue, Mar 17, 2015 at 01:32:44PM +0800, Liu Yuan wrote:
On Tue, Mar 17, 2015 at 01:16:00PM +0800, Liu Yuan
On Tue, Mar 17, 2015 at 02:59:12PM +0900, Hitoshi Mitake wrote:
At Tue, 17 Mar 2015 13:54:16 +0800,
Liu Yuan wrote:
On Tue, Mar 17, 2015 at 01:40:51PM +0800, Liu Yuan wrote:
On Tue, Mar 17, 2015 at 01:32:44PM +0800, Liu Yuan wrote:
On Tue, Mar 17, 2015 at 01:16:00PM +0800, Liu Yuan
On Tue, Mar 17, 2015 at 04:44:46AM +0900, MORITA Kazutaka wrote:
At Mon, 16 Mar 2015 21:13:29 +0800,
Liu Yuan wrote:
How about make 'dog vdi clone --no-share' as the default clone operation?
And
we can add dog vdi clone --share to keep old behavior as optional. By this
manner
On Tue, Mar 17, 2015 at 11:42:01AM +0900, Hitoshi Mitake wrote:
At Tue, 17 Mar 2015 10:03:53 +0800,
Liu Yuan wrote:
On Tue, Mar 17, 2015 at 04:44:46AM +0900, MORITA Kazutaka wrote:
At Mon, 16 Mar 2015 21:13:29 +0800,
Liu Yuan wrote:
How about make 'dog vdi clone --no-share
On Tue, Mar 17, 2015 at 10:03:53AM +0800, Liu Yuan wrote:
On Tue, Mar 17, 2015 at 04:44:46AM +0900, MORITA Kazutaka wrote:
At Mon, 16 Mar 2015 21:13:29 +0800,
Liu Yuan wrote:
How about make 'dog vdi clone --no-share' as the default clone operation?
And
we can add dog vdi clone
On Mon, Mar 16, 2015 at 09:37:45PM +0900, Hitoshi Mitake wrote:
On Mon, Mar 16, 2015 at 4:39 PM, Liu Yuan namei.u...@gmail.com wrote:
On Mon, Mar 16, 2015 at 03:55:43PM +0900, Hitoshi Mitake wrote:
At Mon, 16 Mar 2015 14:48:47 +0800,
Liu Yuan wrote:
On Mon, Mar 16, 2015 at 03:39:17PM
From: Liu Yuan liuy...@cmss.chinamobile.com
This reverts commit b30374cf653c6bf93c91e14fe0954c58e4097637.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
sheep/group.c | 7 ++-
sheep/plain_store.c | 2 +-
sheep/sheep_priv.h | 2 -
sheep/vdi.c | 146
From: Liu Yuan liuy...@cmss.chinamobile.com
This patch fixes following problem:
$ dog vdi create test 100M
$ dog vdi delete test
$ dog vdi create test 200M
$ dog vdi list # expect show test, but nothing shows out.
Which was brought by
*commit f68feab7edc0ded86701a2e902d85616b24942ab
*Author
From: Liu Yuan liuy...@cmss.chinamobile.com
This reverts commit fcb91648b1a1f97c55c2a42fff1ec92b93bfea44.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
include/internal_proto.h | 1 -
sheep/group.c| 3 +--
sheep/ops.c | 7 +++
sheep/plain_store.c
From: Liu Yuan liuy...@cmss.chinamobile.com
This reverts commit 21549a1bd4981fabcc09d062a647162127fe0637.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
sheep/vdi.c | 7 +++
1 file changed, 7 insertions(+)
diff --git a/sheep/vdi.c b/sheep/vdi.c
index dd01a20..af96548 100644
From: Liu Yuan liuy...@cmss.chinamobile.com
This reverts commit 5d005d41862647f6c7cdb7977f2611e61abc104b.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
tests/functional/016.out | 12 ++--
tests/functional/044.out | 12 ++--
tests/functional/046.out | 6 +++---
tests
From: Liu Yuan liuy...@cmss.chinamobile.com
This reverts commit 58f0d6dcdfa370df4456ed08cd32983731fc0c4e.
Signed-off-by: Liu Yuan liuy...@cmss.chinamobile.com
---
sheep/ops.c| 1 -
sheep/sheep_priv.h | 1 -
sheep/vdi.c| 110
From: Liu Yuan liuy...@cmss.chinamobile.com
This patch set mainly revert the patch set that broke the old vdi recycling
algorithm that allow vdi to be recycled if it is deleted, though we keep deleted
inodes in the storage because of vdi allocation algorithm assumes it.
The old vid recycle work
On Mon, Mar 16, 2015 at 09:39:35PM +0900, Hitoshi Mitake wrote:
On Mon, Mar 16, 2015 at 7:42 PM, Liu Yuan namei.u...@gmail.com wrote:
On Mon, Mar 16, 2015 at 03:57:40PM +0900, Hitoshi Mitake wrote:
This patch adds a new option -R to dog cluster format. If user
specifies the option during
On Mon, Mar 16, 2015 at 08:49:34PM +0800, Liu Yuan wrote:
From: Liu Yuan liuy...@cmss.chinamobile.com
This patch fixes following problem:
$ dog vdi create test 100M
$ dog vdi delete test
$ dog vdi create test 200M
$ dog vdi list # expect show test, but nothing shows out.
Which
On Tue, Mar 17, 2015 at 03:55:56AM +, morita...@nttdata.co.jp wrote:
I'm still not sure which code in vdi_lookup() is a problem. The
problem happens even when we disable VID garbage collection?
vdi_lookup() becomes a problem if Hitoshi's patch is enabled after he
fixes a
On Tue, Mar 17, 2015 at 12:08:39PM +0800, redtone wrote:
Many people want do daily snapshot (delete the old one and create a new one
with the save name)
This is not the case we talked about. Your case will work fine. I don't think
you will delete all the snapshots since day 1 very frequently.
On Tue, Mar 17, 2015 at 02:03:03PM +0900, Hitoshi Mitake wrote:
At Tue, 17 Mar 2015 13:58:08 +0900,
Hitoshi Mitake wrote:
At Tue, 17 Mar 2015 13:51:48 +0900,
Hitoshi Mitake wrote:
At Tue, 17 Mar 2015 12:42:35 +0800,
Liu Yuan wrote:
On Mon, Mar 16, 2015 at 08:49:34PM
On Tue, Mar 17, 2015 at 01:16:00PM +0800, Liu Yuan wrote:
On Tue, Mar 17, 2015 at 02:03:03PM +0900, Hitoshi Mitake wrote:
At Tue, 17 Mar 2015 13:58:08 +0900,
Hitoshi Mitake wrote:
At Tue, 17 Mar 2015 13:51:48 +0900,
Hitoshi Mitake wrote:
At Tue, 17 Mar 2015 12:42:35 +0800
On Tue, Mar 17, 2015 at 01:32:44PM +0800, Liu Yuan wrote:
On Tue, Mar 17, 2015 at 01:16:00PM +0800, Liu Yuan wrote:
On Tue, Mar 17, 2015 at 02:03:03PM +0900, Hitoshi Mitake wrote:
At Tue, 17 Mar 2015 13:58:08 +0900,
Hitoshi Mitake wrote:
At Tue, 17 Mar 2015 13:51:48 +0900
On Tue, Mar 17, 2015 at 01:40:51PM +0800, Liu Yuan wrote:
On Tue, Mar 17, 2015 at 01:32:44PM +0800, Liu Yuan wrote:
On Tue, Mar 17, 2015 at 01:16:00PM +0800, Liu Yuan wrote:
On Tue, Mar 17, 2015 at 02:03:03PM +0900, Hitoshi Mitake wrote:
At Tue, 17 Mar 2015 13:58:08 +0900,
Hitoshi
On Tue, Mar 17, 2015 at 01:33:58PM +0900, Hitoshi Mitake wrote:
At Tue, 17 Mar 2015 11:06:34 +0800,
Liu Yuan wrote:
On Tue, Mar 17, 2015 at 11:42:01AM +0900, Hitoshi Mitake wrote:
At Tue, 17 Mar 2015 10:03:53 +0800,
Liu Yuan wrote:
On Tue, Mar 17, 2015 at 04:44:46AM +0900
On Mon, Mar 16, 2015 at 02:36:57PM +0900, Hitoshi Mitake wrote:
At Mon, 16 Mar 2015 10:21:50 +0800,
Liu Yuan wrote:
On Thu, Mar 12, 2015 at 08:14:33PM +0900, Hitoshi Mitake wrote:
At Thu, 12 Mar 2015 14:41:56 +0800,
Liu Yuan wrote:
On Tue, Jan 13, 2015 at 10:37:40AM +0900
On Mon, Mar 16, 2015 at 03:24:31PM +0900, Hitoshi Mitake wrote:
This patch adds a new option -R to dog cluster format. If user
specifies the option during cluster format, the cluster will enable
recycling VID (disabled in default).
Signed-off-by: Hitoshi Mitake mitake.hito...@lab.ntt.co.jp
On Mon, Mar 16, 2015 at 03:55:43PM +0900, Hitoshi Mitake wrote:
At Mon, 16 Mar 2015 14:48:47 +0800,
Liu Yuan wrote:
On Mon, Mar 16, 2015 at 03:39:17PM +0900, Hitoshi Mitake wrote:
At Mon, 16 Mar 2015 14:31:44 +0800,
Liu Yuan wrote:
On Mon, Mar 16, 2015 at 02:08:25PM +0800
On Mon, Mar 16, 2015 at 03:39:17PM +0900, Hitoshi Mitake wrote:
At Mon, 16 Mar 2015 14:31:44 +0800,
Liu Yuan wrote:
On Mon, Mar 16, 2015 at 02:08:25PM +0800, Liu Yuan wrote:
On Mon, Mar 16, 2015 at 02:36:57PM +0900, Hitoshi Mitake wrote:
At Mon, 16 Mar 2015 10:21:50 +0800,
Liu
On Mon, Mar 16, 2015 at 03:54:30PM +0900, Saeki Masaki wrote:
On 2015/03/16 11:06, Liu Yuan wrote:
On Thu, Mar 12, 2015 at 01:50:26PM +0900, Saeki Masaki wrote:
Current sheepdog stores whole objects in single directory like
/var/lib/sheepdog/obj
This mechanism is difficult to handle massive
On Mon, Mar 16, 2015 at 02:08:25PM +0800, Liu Yuan wrote:
On Mon, Mar 16, 2015 at 02:36:57PM +0900, Hitoshi Mitake wrote:
At Mon, 16 Mar 2015 10:21:50 +0800,
Liu Yuan wrote:
On Thu, Mar 12, 2015 at 08:14:33PM +0900, Hitoshi Mitake wrote:
At Thu, 12 Mar 2015 14:41:56 +0800,
Liu
On Mon, Mar 16, 2015 at 03:57:40PM +0900, Hitoshi Mitake wrote:
This patch adds a new option -R to dog cluster format. If user
specifies the option during cluster format, the cluster will enable
recycling VID (disabled in default).
Signed-off-by: Hitoshi Mitake mitake.hito...@lab.ntt.co.jp
On Thu, Mar 12, 2015 at 02:08:13PM +0800, Meng Lingkun wrote:
From: Meng Lingkun mengling...@cmss.chinamobile.com
Add option -c/--cluster to specify the cluster which zk_control
will handle.
Note: Using malloc rather than xmalloc because it can't be linked
under current makefile. Maybe fix
On Thu, Mar 12, 2015 at 03:17:35PM +0300, Vasiliy Tolstov wrote:
Why sheepdog complains about maximum open files limit is too small and
not increase it via
http://man7.org/linux/man-pages/man2/getrlimit.2.html ?
Thanks, we should give it a try, could you post a patch for it and verify
it can
On Thu, Mar 12, 2015 at 08:14:33PM +0900, Hitoshi Mitake wrote:
At Thu, 12 Mar 2015 14:41:56 +0800,
Liu Yuan wrote:
On Tue, Jan 13, 2015 at 10:37:40AM +0900, Hitoshi Mitake wrote:
Current sheepdog never recycles VIDs. But it will cause problems
e.g. VID space exhaustion, too much
On Tue, Jan 13, 2015 at 10:37:40AM +0900, Hitoshi Mitake wrote:
Current sheepdog never recycles VIDs. But it will cause problems
e.g. VID space exhaustion, too much garbage inode objects.
Keeping deleted inode objects is required because living inodes
(snapshots or clones) can point objects
On Thu, Mar 12, 2015 at 08:27:02AM +0100, Valerio Pachera wrote:
2015-03-12 2:51 GMT+01:00 Liu Yuan namei.u...@gmail.com:
install.sh is generated by autoreconf, I guess, controled by configure.ac.
You can verify this by rm install.sh and run ./autogen.sh. So the fix
should go
On Wed, Mar 11, 2015 at 04:05:25PM +0100, Valerio Pachera wrote:
From: Valerio Pachera siri...@gmail.conf
This patch add the package yasm to be installed in both debian 7
and debian 8.
liburcu2 will be insalled instead of liburcu1 in debian 8.
It also fixed an error message that didn't get
On Sun, Mar 08, 2015 at 01:08:33PM +0900, Hitoshi Mitake wrote:
Current sheep cannot handle a case of VID overflow correctly. The main
cause of this problem is invalid usage of fill_vdi_info_range() by
fill_vdi_info() in a case of overflow. This patch fixes this problem.
Could you please
On Mon, Mar 09, 2015 at 07:30:40PM -0700, Alexander Guy wrote:
Common versions of automake have a bug[1] related to enabling
subdir-objects while using variables inside *_SOURCES declarations.
In order to work around these problems with the unit tests, copy
source dependencies locally for
On Tue, Mar 10, 2015 at 12:07:30PM +0300, Vasiliy Tolstov wrote:
Hello. I'm ready to write etcd cluster driver, but i have some absent info.
For example sheperd cluster driver have struct:
static struct cluster_driver cdrv_shepherd = {
.name = shepherd,
.init
1 - 100 of 4133 matches
Mail list logo