Date:	Wed, 22 Jun 2011 09:38:24 +0300
From:	"Amir G." <amir73il@...rs.sourceforge.net>
To:	Andreas Dilger <adilger@...ger.ca>
Cc:	Theodore Tso <tytso@....edu>,
	Ext4 Developers List <linux-ext4@...r.kernel.org>
Subject: Re: [PATCH v1 00/30] Ext4 snapshots

On Tue, Jun 21, 2011 at 6:45 PM, Andreas Dilger <adilger@...ger.ca> wrote:
> On 2011-06-21, at 5:06 AM, "Amir G." <amir73il@...rs.sourceforge.net> wrote:
>> On Tue, Jun 7, 2011 at 6:07 PM,  <amir73il@...rs.sourceforge.net> wrote:
>>>
>>> I am resending the snapshots patch series as per Lukas's request.
>>> This time, the snapshot*.c files have not been omitted, as in
>>> the previous posting.
>>>
>>> The series is still based on ext4 dev branch sometime in the preparation
>>> for 3.0 merge window. It was not yet rebased on 3.0-rc1, so punch holes
>>> changes have not been addressed yet.
>>>
>>> As always, I advocate online review of the patches at:
>>> https://github.com/amir73il/ext4-snapshots/commits/for-ext4-v1
>>> but if you insist on doing it the old way, I won't complain.
>>>
>>>
>>
>> To answer your question about possible diet to snapshot patches,
>> following are some diffstat on groups of patches (by functionality).
>> The diffsstat includes only ext4 core file (excluding new snapshot* files).
>>
>> As you can see, removing the shrink/merge functionality for multiple
>> snapshots, will remove the need for exclude bitmap and reduce the changes
>> to core files by ~300 insertions.
>>
>> There is some advantage of not having to add the exclude bitmap to existing
>> fs, but I think the win is not so big, assuming that we will want to have the
>> shrink/merge functionality eventually.
>
> Wouldn't that also mean if shrink/merge are added later it needs another feature flag and added complexity due to another row in the feature test matrix?
>

Sure, it would add another row in the matrix and we don't want that either.
Shrink/merge capability could be derived from the exclude_bitmap
(compatible) feature.
In fact, without exclude bitmap, shrink/merge can still be done, but
the shrink will
not be as efficient as with exclude_bitmap, but the shrink code works
just the same
(i.e. it doesn't have to check for exclude_bitmap feature).

BTW, I have just learned that ZFS disk space is not reclaimed at all
unless deleting
a complete part of the snapshots history (including the oldest).
This is essentially how ext4 snapshots would work without the
shrink/merge patches
and exclude_bitmap, but it is not very useful for long snapshot
retention policies
(i.e. 1 yearly, 3 monthly, 4 weekly,...).


>
>>
>>> [PATCH v1 01/36] ext4: EXT4 snapshots (Experimental)
>>> [PATCH v1 02/36] ext4: snapshot debugging support
>>
>> Generic stuff.
>>
>> 15 files changed, 122 insertions(+), 1 deletions(-)
>>
>>> [PATCH v1 03/36] ext4: snapshot hooks - inside JBD hooks
>>> [PATCH v1 04/36] ext4: snapshot hooks - block bitmap access
>>> [PATCH v1 05/36] ext4: snapshot hooks - delete blocks
>>> [PATCH v1 06/36] ext4: snapshot hooks - move data blocks
>>> [PATCH v1 07/36] ext4: snapshot hooks - direct I/O
>>> [PATCH v1 08/36] ext4: snapshot hooks - move extent file data blocks
>>
>> Most of the code in this group handles MOW.
>>
>> 7 files changed, 550 insertions(+), 68 deletions(-)
>>
>>> [PATCH v1 09/36] ext4: snapshot file
>>> [PATCH v1 10/36] ext4: snapshot file - read through to block device
>>> [PATCH v1 11/36] ext4: snapshot file - permissions
>>> [PATCH v1 12/36] ext4: snapshot file - store on disk
>>> [PATCH v1 13/36] ext4: snapshot file - increase maximum file size limit to 16TB
>>
>> Implementation of special snapshot file.
>>
>> 7 files changed, 220 insertions(+), 14 deletions(-)
>>
>>> [PATCH v1 14/36] ext4: snapshot block operations
>>> [PATCH v1 15/36] ext4: snapshot block operation - copy blocks to snapshot
>>> [PATCH v1 16/36] ext4: snapshot block operation - move blocks to snapshot
>>> [PATCH v1 17/36] ext4: snapshot block operation - copy block bitmap to snapshot
>>
>> Copy/move to snapshot file operations.
>>
>> 5 files changed, 126 insertions(+), 24 deletions(-)
>>
>>> [PATCH v1 18/36] ext4: snapshot control
>>> [PATCH v1 19/36] ext4: snapshot control - init new snapshot
>>> [PATCH v1 20/36] ext4: snapshot control - fix new snapshot
>>> [PATCH v1 21/36] ext4: snapshot control - reserve disk space for snapshot
>>
>> Mostly new code in ioctl.c.
>>
>> 6 files changed, 171 insertions(+), 3 deletions(-)
>>
>>> [PATCH v1 22/36] ext4: snapshot journaled - increase transaction credits
>>> [PATCH v1 23/36] ext4: snapshot journaled - implement journal_release_buffer()
>>> [PATCH v1 24/36] ext4: snapshot journaled - bypass to save credits
>>> [PATCH v1 25/36] ext4: snapshot journaled - cache last COW tid in journal_head
>>
>> Helper functions to handle extra COW credits.
>>
>> 7 files changed, 284 insertions(+), 31 deletions(-)
>>
>>> [PATCH v1 26/36] ext4: snapshot journaled - trace COW/buffer credits
>>
>> Trace/debug for extra COW credits.
>>
>> 3 files changed, 108 insertions(+), 0 deletions(-)
>>
>>> [PATCH v1 27/36] ext4: snapshot list support
>>> [PATCH v1 28/36] ext4: snapshot list - read through to previous snapshot
>>
>> Not much to gain from dropping snapshot list support...
>>
>> 2 files changed, 2 insertions(+), 0 deletions(-)
>>
>>> [PATCH v1 29/36] ext4: snapshot race conditions - concurrent COW bitmap operations
>>> [PATCH v1 30/36] ext4: snapshot race conditions - concurrent COW operations
>>> [PATCH v1 31/36] ext4: snapshot race conditions - tracked reads
>>
>> We must handle race conditions.
>>
>> 2 files changed, 32 insertions(+), 0 deletions(-)
>>
>>> [PATCH v1 32/36] ext4: snapshot exclude - the exclude bitmap
>>
>> We do not need the exclude bitmap if we do not shrink/merge snapshots.
>>
>> 5 files changed, 226 insertions(+), 3 deletions(-)
>>
>>> [PATCH v1 33/36] ext4: snapshot cleanup
>>> [PATCH v1 34/36] ext4: snapshot cleanup - shrink deleted snapshots
>>> [PATCH v1 35/36] ext4: snapshot cleanup - merge shrunk snapshots
>>
>> We could support multiple snapshots without shrinking/merging deleted snapshot,
>> but that means that only disk space of oldest snapshot is reclaimed on delete.
>>
>> 2 files changed, 74 insertions(+), 22 deletions(-)
>>
>>> [PATCH v1 36/36] ext4: snapshot rocompat - enable rw mount
>>>
>>>  fs/ext4/Kconfig           |   11 +
>>>  fs/ext4/Makefile          |    3 +
>>>  fs/ext4/balloc.c          |  132 +++
>>>  fs/ext4/ext4.h            |  188 ++++-
>>>  fs/ext4/ext4_jbd2.c       |  162 ++++-
>>>  fs/ext4/ext4_jbd2.h       |  266 ++++++-
>>>  fs/ext4/extents.c         |  157 ++++-
>>>  fs/ext4/file.c            |   11 +-
>>>  fs/ext4/ialloc.c          |   19 +-
>>>  fs/ext4/inode.c           |  668 +++++++++++++--
>>>  fs/ext4/ioctl.c           |  120 +++
>>>  fs/ext4/mballoc.c         |  161 ++++-
>>>  fs/ext4/move_extent.c     |    3 +-
>>>  fs/ext4/namei.c           |    9 +
>>>  fs/ext4/resize.c          |   19 +-
>>>  fs/ext4/snapshot.c        | 1000 ++++++++++++++++++++++
>>>  fs/ext4/snapshot.h        |  690 ++++++++++++++++
>>>  fs/ext4/snapshot_buffer.c |  393 +++++++++
>>>  fs/ext4/snapshot_ctl.c    | 2002 +++++++++++++++++++++++++++++++++++++++++++++
>>>  fs/ext4/snapshot_debug.c  |  107 +++
>>>  fs/ext4/snapshot_debug.h  |  105 +++
>>>  fs/ext4/snapshot_inode.c  |  960 ++++++++++++++++++++++
>>>  fs/ext4/super.c           |  157 ++++-
>>>  fs/ext4/xattr.c           |    4 +-
>>>  24 files changed, 7182 insertions(+), 165 deletions(-)
>>>
>>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
>> the body of a message to majordomo@...r.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to majordomo@...r.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
--
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Powered by blists - more mailing lists