You are viewing a plain text version of this content. The canonical link for it is here.

Posted to oak-dev@jackrabbit.apache.org by Michael Dürig <md...@apache.org> on 2014/12/01 13:48:42 UTC

Merging r1642667 to 1.0

Hi,

I propose to merge r1642667 [1] into 1.0.

With OAK-2294 it was discovered that concurrently adding many properties 
to a node might exhaust the 255 segment references limit of the 
SegmentMK. The proposed changes don't fix the problem but mitigate the 
consequences: the affected commit will fail instead of the repository 
becoming corrupted.

If no one object, I'll move forward with the merge later this week.

Michael


[1] http://svn.apache.org/r1642667

Re: Merging r1642667 to 1.0

Posted by Michael Dürig <md...@apache.org>.

Merged at http://svn.apache.org/r1643122

Michael

On 1.12.14 1:50 , Alex Parvulescu wrote:
> +1 even if this is not a solution to the issue, nothing should be able to
> corrupt the repo.
>
> alex
>
> On Mon, Dec 1, 2014 at 1:48 PM, Michael Dürig <md...@apache.org> wrote:
>
>>
>> Hi,
>>
>> I propose to merge r1642667 [1] into 1.0.
>>
>> With OAK-2294 it was discovered that concurrently adding many properties
>> to a node might exhaust the 255 segment references limit of the SegmentMK.
>> The proposed changes don't fix the problem but mitigate the consequences:
>> the affected commit will fail instead of the repository becoming corrupted.
>>
>> If no one object, I'll move forward with the merge later this week.
>>
>> Michael
>>
>>
>> [1] http://svn.apache.org/r1642667
>>
>

Re: Merging r1642667 to 1.0

Posted by Alex Parvulescu <al...@gmail.com>.

+1 even if this is not a solution to the issue, nothing should be able to
corrupt the repo.

alex

On Mon, Dec 1, 2014 at 1:48 PM, Michael Dürig <md...@apache.org> wrote:

>
> Hi,
>
> I propose to merge r1642667 [1] into 1.0.
>
> With OAK-2294 it was discovered that concurrently adding many properties
> to a node might exhaust the 255 segment references limit of the SegmentMK.
> The proposed changes don't fix the problem but mitigate the consequences:
> the affected commit will fail instead of the repository becoming corrupted.
>
> If no one object, I'll move forward with the merge later this week.
>
> Michael
>
>
> [1] http://svn.apache.org/r1642667
>