Hi Craig,

I turned off the write cache on the data sever last night (more accurately, I allocated all the cache to be read cache - that is still active; I also have the option of disabling the cache entirely). This is on the array controller (it is a RAID 5 setup). I still seem to have some delay.

I am now thinking of also trying what someone else suggested, that is, closing the 2 tables in question between "polling" of the queue. This happens every 3 seconds (could go back to 5 or 10 if I had to). Was a little worried that that was a lot of file opening activity but then again, this is 10 years after first implemented so OS's are a whole lot more robust. Any thoughts on that?

Also, not sure what you meant in your sentance below (re the memo fields)? Could you expand?

Thanks,
Albert

>I'm not sure this will solve the original problem of getting the .dbf data, but not memo fields. It's also dependent on what's been cleared from the write cache.
>
>
>>Hi Brandon,
>>
>>Can you explain to me what below does? The RECCOUNT() picks the count up off the header of the file...what does GO RECNO() accomplish?
>>
>>Albert
>>
>>>Try this on applicable tables before your sql-select
>>>
>>>

>>>** Eliminates cached data issues
>>>TRY
>>>	SELECT (lcAlias)
>>>	=RECCOUNT()
>>>	IF !EOF()
>>>		GO RECNO() IN ALIAS()
>>>	ENDIF
>>>CATCH
>>>	** doesn't matter
>>>ENDTRY
>>>