Re: [HACKERS] WIP Patch: Pgbench Serialization and deadlock errors

On Fri, Jun 16, 2017 at 9:18 AM, Alvaro Herrera
<alvherre@2ndquadrant.com> wrote:
> Kevin Grittner wrote:
>> On Thu, Jun 15, 2017 at 2:16 PM, Andres Freund <andres@anarazel.de> wrote:
>> > On 2017-06-14 11:48:25 +0300, Marina Polyakova wrote:
>
>> >> P.S. Does this use case (do not retry transaction with serialization or
>> >> deadlock failure) is most interesting or failed transactions should be
>> >> retried (and how much times if there seems to be no hope of success...)?
>> >
>> > I can't quite parse that sentence, could you restate?
>>
>> The way I read it was that the most interesting solution would retry
>> a transaction from the beginning on a serialization failure or
>> deadlock failure.
>
> As far as I understand her proposal, it is exactly the opposite -- if a
> transaction fails, it is discarded.  And this P.S. note is asking
> whether this is a good idea, or would we prefer that failing
> transactions are retried.
>
> I think it's pretty obvious that transactions that failed with
> some serializability problem should be retried.

+1 for retry with reporting of retry rates

-- 
Thomas Munro
http://www.enterprisedb.com