In-Reply-To: <op.tj43bqe481rjf6@xjens.apus.local>
References: <1165388652.6396.1.camel@panoramix>
	<30BD1C9A-1664-4290-B343-0FB1FEBA7B53@fastcrypt.com>
	<op.tj43bqe481rjf6@xjens.apus.local>
Mime-Version: 1.0 (Apple Message framework v752.3)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <4ED560E4-3484-43BA-99BB-18888D968F74@fastcrypt.com>
Cc: "Joost Kraaijeveld" <J.Kraaijeveld@askesis.nl>,
	pgsql-general@postgresql.org, pgsql-performance@postgresql.org
Content-Transfer-Encoding: 7bit
From: Dave Cramer <pg@fastcrypt.com>
Subject: Re: [PERFORM] Locking in PostgreSQL?
Date: Wed, 6 Dec 2006 08:26:45 -0500
To: Jens Schipkowski <jens.schipkowski@apus.co.at>


On 6-Dec-06, at 8:20 AM, Jens Schipkowski wrote:

> On Wed, 06 Dec 2006 13:29:37 +0100, Dave Cramer <pg@fastcrypt.com>  
> wrote:
>
>> Unless you specifically ask for it postgresql doesn't lock any  
>> rows when you update data.
>>
> Thats not right. UPDATE will force a RowExclusiveLock to rows  
> matching the WHERE clause, or all if no one is specified.
> @Joost Kraaijeveld: Yes, because there is no EntryExclusiveLock or  
> something like that. Roughly you can say, each UPDATE statement  
> iterates through the affected table and locks the WHERE clause  
> matching records (rows) exclusivly to prevent data inconsistancy  
> during the UPDATE. After that your rows will be updated and the  
> lock will be repealed.
> You can see this during an long lasting UPDATE by querying the  
> pg_locks with joined pg_stats_activity (statement must be enabled).

Apparently I've completely misunderstood MVCC then.... My  
understanding is that unless you do a select ... for update then  
update the rows will not be locked .

Dave
>
>> Dave
>> On 6-Dec-06, at 2:04 AM, Joost Kraaijeveld wrote:
>>
>>> Does PostgreSQL lock the entire row in a table if I update only 1
>>> column?
>>>
>>>
>>> --Groeten,
>>>
>>> Joost Kraaijeveld
>>> Askesis B.V.
>>> Molukkenstraat 14
>>> 6524NB Nijmegen
>>> tel: 024-3888063 / 06-51855277
>>> fax: 024-3608416
>>> web: www.askesis.nl
>>>
>>> ---------------------------(end of  
>>> broadcast)---------------------------
>>> TIP 9: In versions below 8.0, the planner will ignore your desire to
>>>        choose an index scan if your joining column's datatypes do  
>>> not
>>>        match
>>>
>>
>>
>> ---------------------------(end of  
>> broadcast)---------------------------
>> TIP 9: In versions below 8.0, the planner will ignore your desire to
>>        choose an index scan if your joining column's datatypes do not
>>        match
>
>
>
> CU,
> Jens
>
> -- 
> **
> Jens Schipkowski
>
> ---------------------------(end of  
> broadcast)---------------------------
> TIP 5: don't forget to increase your free space map settings
>