Excel connect to Access with low performance - excel

Question

My code want to query access very frequently, I use "for" for every row and check a cell's value if existed in access table. But I feel the performance is very bad. Now I'm using adodb.connection to connect to access. Sorry I cannot put code because its not in my hand. Anyone could help me about how to query a table very frequently from excel vba and with fast performance?
EDIT:
For rowNum = 2 To 1000000
'check if title exists,if yes, get ppid, if not, insert one, get ppid, and make relation in r-table
ppID = isTitleExistReturnID(ppTitle)
If ppID = "0" Then
ppID = addPpReturnID(ppTitle, ppDate, ppJournal)
paperAddedCount = paperAddedCount + 1
isPpAdded = True
Else
isPpAdded = False
End If
Next rowNum
Function isTitleExistReturnID(title As String) As String
Dim r As New ADODB.Recordset
sqlstr = "select * from paper where title = '" & title & " '"
'MsgBox sqlstr
dbConnection.Open
r.Open sqlstr, dbConnection, adOpenKeyset, adLockOptimistic, adCmdText
If r.RecordCount < 1 Then
dbConnection.Close
isTitleExistReturnID = "0"
Else
aidi = r.Fields(0).Value
dbConnection.Close
isTitleExistReturnID = aidi
End If
End Function
Function addPpReturnID(title As String, pubDate As String, journaL As String) As String
Dim r As New ADODB.Recordset
sqlstr = "select * from paper where (1=0)"
'MsgBox sqlstr
dbConnection.Open
r.Open sqlstr, dbConnection, adOpenKeyset, adLockOptimistic, adCmdText
r.AddNew
r.Fields(1) = title
r.Fields(2) = pubDate
r.Fields(3) = journaL
r.Update
maxid = CStr(r.Fields(0).Value)
dbConnection.Close
addPpReturnID = maxid
End Function
The above is part of my code:
check if the item is in access table
if YES, return its ID
if NO, add this item and return ID
Do it for more than 100,000 times, very low performance
Any advice will be so appreciated, thanks in advance.

Answer 1

Any RBAR (row by agonizing row) approach has the potential to be a performance challenge --- that's why it's called agonizing.
And yours is definitely RBAR because you do stuff separately for each of a million (For rowNum = 2 To 1000000) spreadsheet rows. Compounding the problem is the fact that you're opening and closing ADODB objects (connections and recordsets) at least one, but sometimes twice, for each of those rows.
Try to find a set-based approach instead. For example, if you can drive this operation from Access ...
Create a link to the spreadsheet.
Create an "unmatched" query (there is a query wizard to guide you) to select those spreadsheet rows whose ppTitle does not exist in the Access destination table.
Create an "append" query to add those unmatched rows to the Access table.
I don't know whether that outline is a suitable fit for your situation or whether it's close enough that you can adapt it. But the more important point is to find a set-based method instead of RBAR.