Bob,
Nope, it doesn't, not at all. Oddly as I continue to test,it turns out that Markus's suggestion that I click the mouse, while waiting for it to download, indeed has A DRAMATIC effect on performance. Markus had suggested clicking once. I find that if I just continually move the mouse over my desktop, I can get it to finish in 20 secs in stead of 3 minutes. Bizarre but true.
Isn't there a guru out there who can explain this odd behaviour? In the meantime, until someone tells me otherwise, I need to conclude that even though I was under the impression that remote views are just a 'simple' wrapper around spt, I need to conclude that there are big differences in the implementation, and that I will definitely be giving up remote views entirely in favor of updatable spt cursors.
>USE MyRemoteView NOUPDATE
>
>See if that comes in faster.
>
>BOb
>
>
>>Mike,
>>
>>There are some conditions that I am working on where I might be able to limit the result set. Even so, at this point here is my main issue:
>>
>>As I said previously - here are the stats on retrieval:
>>
>>using remote view - open view with all 85K, look at status bar - recpointer/reccount() - takes 3 minutes to download all records.
>>
>>lnresult=sqlexe(lnhandle,'select * from sostrs','cursostrs')
>>takes only 6 seconds for command to finish executing, all rows returned
>>I can move record pointer to bottom, middle etc, execute a replace, no problem.
>>
>>Why such a tremendous difference when remote view is just a 'wrapper' around SPT? People I have spoken with have described a 50% difference in remote view vs spt. I am seeing orders of magnitude. Am I overlooking some property setting, or something simple - like 'SET GOFAST ON' (just kidding)
>>
>>Thanks for your help!
>>
>>>>I have a sql server table with 85K rows, a sales order line item file. Unfortunately for this application, it is difficult to come up with parameter conditions to limit the # of rows to retrieve - it makes more sense to download the whole file for the application I am working with.
>>>
>>>I highly suggest that you find something to limit the amount of data that you will be retrieving. If you don't, you will be creating performance issues. What's a user going to do with 85,000 rows? How would they go about finding a specific set of rows? Can you not force some kind of search specs such as a date range, customer number, employee number, country code, something?!?!
>>>
>>>-Mike
Précédent
Suivant
Répondre
Voir le fil de ce thread
Voir le fil de ce thread à partir de ce message seulement
Voir tous les messages de ce thread
Voir tous les messages de ce thread à partir de ce message seulement