|
Home > Archive > SQL Anywhere ultralite > August 2005 > downloading big amount of values
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
downloading big amount of values
|
|
|
| moblink 9.0.2
ultralite 9.0.1
Hello there!
I need to download 71,940 values from Oracle so i devided it
to two tables. one for the values that smaller then 65,000
and the second for the values that bigger then 65,000
(include 65,000).
this two tables are a part of db and i'm using publication
to download only this two tables.
now the problem is ut takes about 10 minutes and sometimes
even more to download the values.
when i check the mobilink log file it says that it dowloaded
the values in two 1 minute but still i don't get download
succeeded only after 10 minutes.
here is my download cod (c#):
try
{
ConnUp.SyncParms.PublicationMask = 4;
ConnUp.SyncParms.Stream = StreamType.TCPIP;
ConnUp.SyncParms.SendColumnNames = true;
//ConnUp.SyncParms.UploadOnly = true;
ConnUp.SyncParms.DownloadOnly = true;
ConnUp.SyncParms.Version = "download";
ConnUp.SyncParms.UserName = Custom.MOBILINK_USER
+ "_u";
ConnUp.SyncParms.Password = Custom.MOBILINK_PASS;
ConnUp.SyncParms.StreamParms = Custom.NEHIM_IP;
ConnUp.Synchronize();
}
catch( SQLException err )
{
MessageBox.Show( "syncException 1: " + err.Message + "
sqlcode=" + err.ErrorCode);
}
is there a reson for this long time??
have a good day,
Omri Ziv
| |
| Greg Fenton 2005-08-15, 3:23 am |
| Omri wrote:
> moblink 9.0.2
> ultralite 9.0.1
> Hello there!
> I need to download 71,940 values from Oracle so i devided it
[...]
> now the problem is ut takes about 10 minutes and sometimes
> even more to download the values.
> when i check the mobilink log file it says that it dowloaded
> the values in two 1 minute but still i don't get download
> succeeded only after 10 minutes.
The ML server builds the download stream on a server-class machine and
hands the download responsibilities off to a "worker thread", so it
considers its work done in a much shorter amount of time (1 minute).
You are downloading a large amount of data and asking that a handheld
device apply 70000+ rows. Those devices are quite slow at applying the
download stream, and this is where the 10 minutes is coming from.
greg.fenton
--
Greg Fenton
Consultant, Solution Services, iAnywhere Solutions
--------
Visit the iAnywhere Solutions Developer Community
Whitepapers, TechDocs, Downloads
http://www.ianywhere.com/developer/
| |
| Tom Slee 2005-08-16, 1:25 pm |
| Greg Fenton wrote:
> Omri wrote:
>
>
>
> [...]
>
>
>
> The ML server builds the download stream on a server-class machine and
> hands the download responsibilities off to a "worker thread", so it
> considers its work done in a much shorter amount of time (1 minute).
>
> You are downloading a large amount of data and asking that a handheld
> device apply 70000+ rows. Those devices are quite slow at applying the
> download stream, and this is where the 10 minutes is coming from.
>
> greg.fenton
Just curious -- how many columns do you have and how heavily is the
mobile database indexed?
Tom Slee
| |
|
| > Greg Fenton wrote:
[color=darkred]
>
> Just curious -- how many columns do you have and how
> heavily is the mobile database indexed?
>
> Tom Slee
Hi Tom,
i have only two columns one for row_num and the other is
prod_car, nothing too much heavy.
regards
Omri Ziv
| |
|
| > >
> Just curious -- how many columns do you have and how
> heavily is the mobile database indexed?
> Tom Slee
Hi Tom,
i have only two columns one for row_num and the other is
prod_car, nothing too much heavy.
regards
Omri Ziv
| |
| Tom Slee 2005-08-19, 8:23 pm |
| Omri wrote:
>
>
> Hi Tom,
> i have only two columns one for row_num and the other is
> prod_car, nothing too much heavy.
>
> regards
> Omri Ziv
On a fairly recent iPAQ or something, I would expect to be able to
insert about 10,000 rows per minute for a small table like this, which
translates into about 7 minutes for your synchronization. So I would
certainly expect a fairly long wait to insert this volume of data. I
wonder if there is a way of identifying what information is needed on
device at a finer granularity. What kind of device are you using?
Tom
| |
|
| > Omri wrote:
[color=darkred]
> if there is a way of identifying what information is
> needed on device at a finer granularity. What kind of
> device are you using?
>
> Tom
Hi Tom,
Unfortunately all the information is needed.
i'm using i-mate.
i found a primitive solution because rthe information is
changing one in two weeks i created an Independent program
tha create the udb file and i just insert it into the
devices.
and i save a lot of working time.
it's a bit primitive but worth the time.
Have a good day
Omri
| |
| Tom Slee 2005-08-29, 11:23 am |
| Omri wrote:
>
>
>
> Hi Tom,
> Unfortunately all the information is needed.
> i'm using i-mate.
> i found a primitive solution because rthe information is
> changing one in two weeks i created an Independent program
> tha create the udb file and i just insert it into the
> devices.
> and i save a lot of working time.
> it's a bit primitive but worth the time.
>
> Have a good day
> Omri
Nothing wrong with primitive if it works. Thanks for letting us know.
Tom
|
|
|
|
|