Re: [Wikitech-l] cleaning database of spam

2013-02-26 Thread Jay Ashworth
- Original Message -
> From: "Petr Bena" 

> You meant innodb_file_per_table

Yes; I forgot the exact name, and tried (apparently unsuccessfully) to 
make that look as little like an exact parameter as possible.

Happily, the OP runs that way anyway.

Cheers,
-- jra
-- 
Jay R. Ashworth  Baylink   j...@baylink.com
Designer The Things I Think   RFC 2100
Ashworth & Associates http://baylink.pitas.com 2000 Land Rover DII
St Petersburg FL USA   #natog  +1 727 647 1274

___
Wikitech-l mailing list
Wikitech-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikitech-l

Re: [Wikitech-l] cleaning database of spam

2013-02-26 Thread Petr Bena
You meant innodb_file_per_table

On Tue, Feb 26, 2013 at 5:39 PM, Petr Bena  wrote:
> yes that's what I do :>
>
> On Tue, Feb 26, 2013 at 4:39 PM, Jay Ashworth  wrote:
>> - Original Message -
>>> From: "Platonides" 
>>
>>> > What is exact procedure of properly removing page from database so
>>> > that it doesn't break anything? What needs to be deleted and in
>>> > which order?
>>>
>>> maintenance/deleteArchivedRevisions.php permanently removes the
>>> content of deleted pages from the db.
>>>
>>> For removing those users, see
>>> http://www.mediawiki.org/wiki/Extension:User_Merge_and_Delete
>>>
>>> Also remember that due to the way mysql works, it may not release
>>> those 20GB back to the filesystem.
>>
>> In particular, to get anything for your trouble, you will probably need
>> to dump the database, drop it, shut down MySQL and switch it to innodb
>> tablespace-per-file, turn it back on, and then reload the dump, as I
>> recently had to.
>>
>> This way, at least, once you clean it up, you can do the same dump and
>> reload procedure on only one table, not the whole shootin' match.
>>
>> Cheers,
>> -- jra
>> --
>> Jay R. Ashworth  Baylink   
>> j...@baylink.com
>> Designer The Things I Think   RFC 
>> 2100
>> Ashworth & Associates http://baylink.pitas.com 2000 Land Rover 
>> DII
>> St Petersburg FL USA   #natog  +1 727 647 
>> 1274
>>
>> ___
>> Wikitech-l mailing list
>> Wikitech-l@lists.wikimedia.org
>> https://lists.wikimedia.org/mailman/listinfo/wikitech-l

___
Wikitech-l mailing list
Wikitech-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikitech-l

Re: [Wikitech-l] cleaning database of spam

2013-02-26 Thread Petr Bena
yes that's what I do :>

On Tue, Feb 26, 2013 at 4:39 PM, Jay Ashworth  wrote:
> - Original Message -
>> From: "Platonides" 
>
>> > What is exact procedure of properly removing page from database so
>> > that it doesn't break anything? What needs to be deleted and in
>> > which order?
>>
>> maintenance/deleteArchivedRevisions.php permanently removes the
>> content of deleted pages from the db.
>>
>> For removing those users, see
>> http://www.mediawiki.org/wiki/Extension:User_Merge_and_Delete
>>
>> Also remember that due to the way mysql works, it may not release
>> those 20GB back to the filesystem.
>
> In particular, to get anything for your trouble, you will probably need
> to dump the database, drop it, shut down MySQL and switch it to innodb
> tablespace-per-file, turn it back on, and then reload the dump, as I
> recently had to.
>
> This way, at least, once you clean it up, you can do the same dump and
> reload procedure on only one table, not the whole shootin' match.
>
> Cheers,
> -- jra
> --
> Jay R. Ashworth  Baylink   
> j...@baylink.com
> Designer The Things I Think   RFC 2100
> Ashworth & Associates http://baylink.pitas.com 2000 Land Rover DII
> St Petersburg FL USA   #natog  +1 727 647 1274
>
> ___
> Wikitech-l mailing list
> Wikitech-l@lists.wikimedia.org
> https://lists.wikimedia.org/mailman/listinfo/wikitech-l

___
Wikitech-l mailing list
Wikitech-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikitech-l

Re: [Wikitech-l] cleaning database of spam

2013-02-26 Thread Jay Ashworth
- Original Message -
> From: "Platonides" 

> > What is exact procedure of properly removing page from database so
> > that it doesn't break anything? What needs to be deleted and in
> > which order?
> 
> maintenance/deleteArchivedRevisions.php permanently removes the
> content of deleted pages from the db.
> 
> For removing those users, see
> http://www.mediawiki.org/wiki/Extension:User_Merge_and_Delete
> 
> Also remember that due to the way mysql works, it may not release
> those 20GB back to the filesystem.

In particular, to get anything for your trouble, you will probably need
to dump the database, drop it, shut down MySQL and switch it to innodb
tablespace-per-file, turn it back on, and then reload the dump, as I 
recently had to.

This way, at least, once you clean it up, you can do the same dump and 
reload procedure on only one table, not the whole shootin' match.

Cheers,
-- jra
-- 
Jay R. Ashworth  Baylink   j...@baylink.com
Designer The Things I Think   RFC 2100
Ashworth & Associates http://baylink.pitas.com 2000 Land Rover DII
St Petersburg FL USA   #natog  +1 727 647 1274

___
Wikitech-l mailing list
Wikitech-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikitech-l

Re: [Wikitech-l] cleaning database of spam

2013-02-26 Thread Petr Bena
but it will stop incrementing the datafile

On Tue, Feb 26, 2013 at 12:40 PM, Platonides  wrote:
> On 26/02/13 11:57, Petr Bena wrote:
>> Hi, this is more related to mediawiki rather than wikimedia, but this
>> list is being watched a bit more I guess.
>>
>> Is there any extension that allows permanent removal of deleted pages
>> (or eventually selected deleted pages) from database and removal of
>> blocked users from database?
>>
>> Imagine you have a mediawiki wiki that has 20 gb database, where
>> 19.99gb of database is spam and indefinitely blocked users. I think
>> lot of wikis has this problem, making extension to deal with this
>> would be useful for many small wikis.
>>
>> What is exact procedure of properly removing page from database so
>> that it doesn't break anything? What needs to be deleted and in which
>> order?
>
> maintenance/deleteArchivedRevisions.php permanently removes the content
> of deleted pages from the db.
>
> For removing those users, see
> http://www.mediawiki.org/wiki/Extension:User_Merge_and_Delete
>
> Also remember that due to the way mysql works, it may not release those
> 20GB back to the filesystem.
>
>
> ___
> Wikitech-l mailing list
> Wikitech-l@lists.wikimedia.org
> https://lists.wikimedia.org/mailman/listinfo/wikitech-l

___
Wikitech-l mailing list
Wikitech-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikitech-l

Re: [Wikitech-l] cleaning database of spam

2013-02-26 Thread Platonides
On 26/02/13 11:57, Petr Bena wrote:
> Hi, this is more related to mediawiki rather than wikimedia, but this
> list is being watched a bit more I guess.
> 
> Is there any extension that allows permanent removal of deleted pages
> (or eventually selected deleted pages) from database and removal of
> blocked users from database?
> 
> Imagine you have a mediawiki wiki that has 20 gb database, where
> 19.99gb of database is spam and indefinitely blocked users. I think
> lot of wikis has this problem, making extension to deal with this
> would be useful for many small wikis.
> 
> What is exact procedure of properly removing page from database so
> that it doesn't break anything? What needs to be deleted and in which
> order?

maintenance/deleteArchivedRevisions.php permanently removes the content
of deleted pages from the db.

For removing those users, see
http://www.mediawiki.org/wiki/Extension:User_Merge_and_Delete

Also remember that due to the way mysql works, it may not release those
20GB back to the filesystem.


___
Wikitech-l mailing list
Wikitech-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikitech-l

[Wikitech-l] cleaning database of spam

2013-02-26 Thread Petr Bena
Hi, this is more related to mediawiki rather than wikimedia, but this
list is being watched a bit more I guess.

Is there any extension that allows permanent removal of deleted pages
(or eventually selected deleted pages) from database and removal of
blocked users from database?

Imagine you have a mediawiki wiki that has 20 gb database, where
19.99gb of database is spam and indefinitely blocked users. I think
lot of wikis has this problem, making extension to deal with this
would be useful for many small wikis.

What is exact procedure of properly removing page from database so
that it doesn't break anything? What needs to be deleted and in which
order?

___
Wikitech-l mailing list
Wikitech-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikitech-l