Re: cleanup temporary files after crash

On 3/17/21 2:34 AM, Euler Taveira wrote:
> On Sun, Mar 14, 2021, at 11:01 PM, Thomas Munro wrote:
>> On Wed, Mar 10, 2021 at 1:31 AM Michael Paquier <michael@paquier.xyz
>> <mailto:michael@paquier.xyz>> wrote:
>> > On Tue, Mar 09, 2021 at 02:28:43AM +0100, Tomas Vondra wrote:
>> > > Let's move this patch forward. Based on the responses, I agree the
>> > > default behavior should be to remove the temp files, and I think we
>> > > should have the GUC (on the off chance that someone wants to preserve
>> > > the temporary files for debugging or whatever other reason).
>> >
>> > Thanks for taking care of this.  I am having some second-thoughts
>> > about changing this behavior by default, still that's much more useful
>> > this way.
>>
>> +1 for having it on by default.
>>
>> I was also just looking at this patch and came here to say LGTM except
>> for two cosmetic things, below.
> Thanks for taking a look at this patch. I'm not sure Tomas is preparing
> a new
> patch that includes the suggested modifications but I decided to do it. This
> new version has the new GUC name (using "remove"). I also replaced "cleanup"
> with "remove" in the all the remain places. As pointed by Thomas, I reworded
> the paragraph that describes the GUC moving the default information to the
> beginning of the sentence. I also added the "literal" as suggested by
> Michael.
> The postgresql.conf.sample was fixed. The tests was slightly modified. I
> reworded some comments and added a hack to avoid breaking the temporary file
> test in slow machines. A usleep() after sending the query provides some time
> for the query to create the temporary file. I used an arbitrarily sleep
> (10ms)
> that seems to be sufficient.
> 

Thanks. Pushed with some minor changes to docs wording.


regards

-- 
Tomas Vondra
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company