|From:||Etsuro Fujita <fujita(dot)etsuro(at)lab(dot)ntt(dot)co(dot)jp>|
|To:||Ashutosh Bapat <ashutosh(dot)bapat(at)enterprisedb(dot)com>|
|Cc:||Rushabh Lathia <rushabh(dot)lathia(at)gmail(dot)com>, pgsql-hackers <pgsql-hackers(at)postgresql(dot)org>|
|Subject:||Re: Push down more UPDATEs/DELETEs in postgres_fdw|
|Views:||Raw Message | Whole Thread | Download mbox|
On 2016/11/16 16:38, Etsuro Fujita wrote:
> On 2016/11/16 13:10, Ashutosh Bapat wrote:
>> I don't see any reason why DML/UPDATE pushdown should depend upon
>> subquery deparsing or least PHV patch. Combined together they can help
>> in more cases, but without those patches, we will be able to push-down
>> more stuff. Probably, we should just restrict push-down only for the
>> cases when above patches are not needed. That makes reviews easy. Once
>> those patches get committed, we may add more functionality depending
>> upon the status of this patch. Does that make sense?
> OK, I'll extract from the patch the minimal part that wouldn't depend on
> the two patches.
Here is a patch for that. Todo items are: (1) add more comments and (2)
add more regression tests. I'll do that in the next version.
|Next Message||Joshua Drake||2016-11-18 03:43:21||Re: Mail thread references in commits|
|Previous Message||Craig Ringer||2016-11-18 03:08:20||Re: Patch: Implement failover on libpq connect level.|