PEP 249 (database api) -- executemany() with iterable?

Lawrence D'Oliveiro ldo at geek-central.gen.new_zealand
Sat Oct 16 03:10:08 CEST 2010


In message
<ecdbf6b3-3f99-4b59-b7f8-85bd22f97cac at w9g2000prc.googlegroups.com>, Steve 
Howell wrote:

> On Oct 13, 8:32 pm, Lawrence D'Oliveiro <ldo at geek-central.gen.new_zealand>
> wrote:
>
>> In message
>> <d2451907-c0d2-4571-b3e1-1e4d4f66a... at a7g2000prb.googlegroups.com>, Steve
>> Howell wrote:
>>
>>> Bulk-load strategies usually solve one or more of these problems:
>>
>>> network latency
>>
>> That’s not an issue. This is a bulk operation, after all.
>>
>>> index maintenance during the upload
>>
>> There are usually options to temporarily turn this off.
>>
>>> parsing of the SQL
>>
>> Not usually a problem, as far as I can tell.
>>
>>> reallocating storage
>>
>> You mean for thr SQL? Again, not usually a problem, as far as I can tell.
> 
> If you are avoiding network latency and turning off indexes, then you
> are using some kind of a bulk-load strategy.

I thought we were talking about “load data” versus using simple variants of 
“insert”. At least, that’s what I was talking about. What were you talking 
about?

> If you are not concerned about parsing costs or storage churn ...

What is this “storage churn” thing, and why are you raising it now?



More information about the Python-list mailing list