Hi all.

>From a deeper reading of the related posts, starting from this one, it 
seems that cursor-semantics are definitely not allowed by NHibernate:

https://groups.google.com/d/topic/nhusers/1toBHtY5y1s/discussion

Cheers,
Marcello.

Il giorno martedì 5 marzo 2013 10:44:44 UTC+1, Marcello Esposito ha scritto:
>
> Hi all.
>
> One more element.
>
> I tried to use Ayende's 
> approach<http://ayende.com/blog/4548/nhibernate-streaming-large-result-sets>with
>  a MySql5 connection (Mysql.Data version 6.5.4.0). The result is the 
> same.
> This time I also used a network sniffer.
>
> The query is raised against the DB.
> The whole result set is streamed.
> Only after the ActionableList::Add method i raised for the first time.
>
> Perhaps am I missing to set some configuration option?
>
> Thanks,
> Marcello.
>
> Il giorno martedì 26 febbraio 2013 12:04:08 UTC+1, Marcello Esposito ha 
> scritto:
>>
>> I just point that my npgsql driver is the 2.0 version.
>> Streaming works well and I successfully tried it with a "plain" 
>> command.ExecuteReader() on a large resultset.
>>
>> Cheers,
>> Marcello.
>>
>> Il giorno martedì 26 febbraio 2013 11:45:48 UTC+1, Marcello Esposito ha 
>> scritto:
>>>
>>> > Actually, it is definitely what I needed.
>>>
>>> ...but it seems not to work.
>>>
>>> My steps are:
>>>
>>>    - I raise the query with a limit of 10.000 records 
>>> (session.CreateQuery("from 
>>>    Entity").SetMaxResults(10000).List(myActionableList))
>>>    - server hard disk starts working
>>>    - network starts measuring a constant traffic for a while
>>>    - during this while, a breakpoint I put over IList::Add() is not 
>>>    triggered
>>>    - about 20 seconds elapse...
>>>    - ...suddenly network load stops
>>>    - breakpoint is triggered
>>>    - releasing the breakpoint all data are written in less than a 
>>>    second (in the while the network is idle and burst-less)
>>>
>>> Request for all the records (about ten millions) doesn't starts writing 
>>> for a long time.
>>>
>>> It seems that streaming is not happening and memory is loaded by the 
>>> full result set before data processing.
>>>
>>> Don't know if it is important: I'm using NH3.3.1.4000 with Postgres as a 
>>> database.
>>>
>>> Has someone successfully tried this approach?
>>>
>>> Thanks a lot and sorry for boring you,
>>> Marcello.
>>>
>>>>

-- 
You received this message because you are subscribed to the Google Groups 
"nhusers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/nhusers?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.


Reply via email to