Jim,

Here's how I do it.

Everything has a unique ID and is date/time stamped upon entry or 
modification. There is another table that tracks data dumps by the 
date/time of the send. In my case there are several different 
categories that might be sent to more than one recipient. Among 
other things I track how many records were sent, if financial what 
the total $$ were, what the max date/time of the record sent was, 
what time time the transfer took place, and a count of how many 
updates have been done for the day. Again, in my case, this has 
been valuable in arguing with the recipient's techs when there is a 
disparity in the data.

The user clicks a menu selection to initiate the whole thing. The 
system compares the last max date/time against the DB, 
assembles the data files, and shells to an FTP script. The user is 
prompted to confirm the FTP happened w/o error and, if so, another 
entry is inserted into the tracking table.  In some cases I also track 
what records (by ID) were sent on what date/time (John Doe's 
record was sent on the First, Tenth, and Fifteenth of last month as 
opposed to only knowing the _last_ time it was sent).

The transferred files remain on the hard drive "just in case"... and 
the file naming convention is UserIDMMDD.c00 [MM=Month, DD = 
Day, c=data category,00 count for the day]

This has been going on at several sites for a couple years now 
(thousands of updates) w/out a hitch. Hope this helps

Ben Petersen


On 3 Jun 2002, at 13:16, Jim Limburg wrote:

> G-day all
> 
> I was wondering. I want to know how most of ya-all handle
> this type of situation. We have a 2 databases we need to 
> keep in sync (Actually same SCHEMA/db/tables/etc). To do this 
> we unload the data at one place using the OUTPUT _to filename_
> using the UNLOAD DATA command into a file and then FTP the 
> file to another location and RUN _the file_ to load it back 
> into the same database/table structure at the second location.
> 
> What happens is we have two users involved in this process.
> One keys in the data, and when she is satsified she prints
> a hardcopy, and this marks the data (one column) for the effected
> records for them to be sent. Then the first user contacts a more 
> experienced user who then runs a routine to UNLOAD the data into 
> files and FTP's them. 
> 
> What happens is that the user who is suppose to send them
> sometimes goofs, or when second user is gone someone else does 
> this wrong or ???? and the UNLOAD sequence gets run more than 
> once creating a file with no records overwriting the first file 
> before it get's FTP'd.
> 
> I am kinda thinking along the lines of a unique file naming schema,
> but this presents a real pain in coding a way to look for what has
> been loaded and what has not, and looking for file names or at 
> least it seems so. I have also wondered about loading and unloading 
> from different tables and figuring out some kind of tracking method, 
> but have not done a lot of research in this area yet.
> 
> I was wondering what methods are used by you all with great success?
> 
> Thanks
> Jim Limburg
> 
> __________________________________________________
> Do You Yahoo!?
> Yahoo! - Official partner of 2002 FIFA World Cup
> http://fifaworldcup.yahoo.com
> ================================================
> TO SEE MESSAGE POSTING GUIDELINES:
> Send a plain text email to [EMAIL PROTECTED]
> In the message body, put just two words: INTRO rbase-l
> ================================================
> TO UNSUBSCRIBE: send a plain text email to [EMAIL PROTECTED]
> In the message body, put just two words: UNSUBSCRIBE rbase-l
> ================================================
> TO SEARCH ARCHIVES:
> http://www.mail-archive.com/rbase-l%40sonetmail.com/
> 


================================================
TO SEE MESSAGE POSTING GUIDELINES:
Send a plain text email to [EMAIL PROTECTED]
In the message body, put just two words: INTRO rbase-l
================================================
TO UNSUBSCRIBE: send a plain text email to [EMAIL PROTECTED]
In the message body, put just two words: UNSUBSCRIBE rbase-l
================================================
TO SEARCH ARCHIVES:
http://www.mail-archive.com/rbase-l%40sonetmail.com/

Reply via email to