Increase page size for SyncData DB from 4K to maximum supported 32K

Increase page size for SyncData DB from 4K to maximum
supported 32K using finch experiment.

BUG=464063
Committed: https://crrev.com/5e541eb34b7e1944b8086705a18e86c892cad4a6
Cr-Commit-Position: refs/heads/master@{#321529}

[Gang Wu, 2015-03-16 23:11:48]

gangwu@chromium.org changed reviewers:
+ stanisc@chromium.org

[Gang Wu, 2015-03-16 23:14:24]

Please take a look.

[stanisc, 2015-03-17 16:02:41]

Looks good overall. It would be good to add a unit test that would ensure that
the DB can still get opened/loaded after upgrading to 32K page size.

https://codereview.chromium.org/1016563005/diff/1/sync/syncable/on_disk_direc...
File sync/syncable/on_disk_directory_backing_store.cc (right):

https://codereview.chromium.org/1016563005/diff/1/sync/syncable/on_disk_direc...
sync/syncable/on_disk_directory_backing_store.cc:92: if
(IsSyncBackingDatabase32KEnabled()) {
This code fragment is repeated 4 times. Consider making it a function in the
base class.
Or I think it should be enough to simply place this code once in the base class
constructor.

[Gang Wu, 2015-03-17 23:52:47]

Added unit test for increase page size.

https://codereview.chromium.org/1016563005/diff/1/sync/syncable/on_disk_direc...
File sync/syncable/on_disk_directory_backing_store.cc (right):

https://codereview.chromium.org/1016563005/diff/1/sync/syncable/on_disk_direc...
sync/syncable/on_disk_directory_backing_store.cc:92: if
(IsSyncBackingDatabase32KEnabled()) {
On 2015/03/17 16:02:41, stanisc wrote:
> This code fragment is repeated 4 times. Consider making it a function in the
> base class.
> Or I think it should be enough to simply place this code once in the base
class
> constructor.

Done.

[Gang Wu, 2015-03-18 03:11:38]

gangwu@chromium.org changed reviewers:
+ asvitkine@chromium.org

[Gang Wu, 2015-03-18 06:01:33]

gangwu@chromium.org changed reviewers:
+ yiyaoliu@chromium.org

[stanisc, 2015-03-18 15:10:45]

ltgm with a couple of nits

https://codereview.chromium.org/1016563005/diff/20001/sync/syncable/directory...
File sync/syncable/directory_backing_store.cc (right):

https://codereview.chromium.org/1016563005/diff/20001/sync/syncable/directory...
sync/syncable/directory_backing_store.cc:180: if
(IsSyncBackingDatabase32KEnabled()) {
Consider ternary "? :" operator here instead of "if else"

https://codereview.chromium.org/1016563005/diff/20001/sync/syncable/directory...
File sync/syncable/directory_backing_store.h (right):

https://codereview.chromium.org/1016563005/diff/20001/sync/syncable/directory...
sync/syncable/directory_backing_store.h:155: int databasePageSize;
Member field should have an underscore at the end.

[stanisc, 2015-03-18 15:11:24]

On 2015/03/18 15:10:45, stanisc wrote:
> ltgm with a couple of nits
> 
>
https://codereview.chromium.org/1016563005/diff/20001/sync/syncable/directory...
> File sync/syncable/directory_backing_store.cc (right):
> 
>
https://codereview.chromium.org/1016563005/diff/20001/sync/syncable/directory...
> sync/syncable/directory_backing_store.cc:180: if
> (IsSyncBackingDatabase32KEnabled()) {
> Consider ternary "? :" operator here instead of "if else"
> 
>
https://codereview.chromium.org/1016563005/diff/20001/sync/syncable/directory...
> File sync/syncable/directory_backing_store.h (right):
> 
>
https://codereview.chromium.org/1016563005/diff/20001/sync/syncable/directory...
> sync/syncable/directory_backing_store.h:155: int databasePageSize;
> Member field should have an underscore at the end.

I meant lgtm.

[Gang Wu, 2015-03-18 23:52:17]

Patchset #3 (id:40001) has been deleted

[Gang Wu, 2015-03-19 04:58:20]

Patchset #3 (id:60001) has been deleted

[Gang Wu, 2015-03-19 05:00:02]

Patchset #3 (id:80001) has been deleted

[Gang Wu, 2015-03-19 22:43:07]

The CQ bit was checked by gangwu@chromium.org

[Gang Wu, 2015-03-19 22:43:08]

The patchset sent to the CQ was uploaded after l-g-t-m from stanisc@chromium.org
Link to the patchset: https://codereview.chromium.org/1016563005/#ps100001
(title: " ")

[commit-bot: I haz the power, 2015-03-19 22:43:15]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1016563005/100001

[Gang Wu, 2015-03-20 05:22:18]

The CQ bit was checked by gangwu@chromium.org

[Gang Wu, 2015-03-20 05:22:18]

The patchset sent to the CQ was uploaded after l-g-t-m from stanisc@chromium.org
Link to the patchset: https://codereview.chromium.org/1016563005/#ps120001
(title: " ")

[commit-bot: I haz the power, 2015-03-20 05:22:44]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1016563005/120001

[commit-bot: I haz the power, 2015-03-20 07:02:25]

Committed patchset #4 (id:120001)

[commit-bot: I haz the power, 2015-03-20 07:03:14]

Patchset 4 (id:??) landed as
https://crrev.com/5e541eb34b7e1944b8086705a18e86c892cad4a6
Cr-Commit-Position: refs/heads/master@{#321529}

[Gang Wu, 2015-05-21 18:38:17]

https://codereview.chromium.org/1016563005/diff/20001/sync/syncable/directory...
File sync/syncable/directory_backing_store.cc (right):

https://codereview.chromium.org/1016563005/diff/20001/sync/syncable/directory...
sync/syncable/directory_backing_store.cc:180: if
(IsSyncBackingDatabase32KEnabled()) {
On 2015/03/18 15:10:45, stanisc wrote:
> Consider ternary "? :" operator here instead of "if else" 

Done.

https://codereview.chromium.org/1016563005/diff/20001/sync/syncable/directory...
File sync/syncable/directory_backing_store.h (right):

https://codereview.chromium.org/1016563005/diff/20001/sync/syncable/directory...
sync/syncable/directory_backing_store.h:155: int databasePageSize;
On 2015/03/18 15:10:45, stanisc wrote:
> Member field should have an underscore at the end.

Done.