Database

Re: database Performance......

30-Jun-09 0:26

MS Access can support even 2 million of records.

But considering other factors like

Feature SQL Server 2000 Access 2000
database size 1,048,516 TB 2 GB plus
linked tables size
objects in a database 2,147,483,647 32,768
user name length 128 20
password length 128 14
table name length 128 64
column name length 128 64
index name length 128 64
Number of concurrent users limited 255
columns per table 1024 255
table size limited 1 GB
number of indexes in a table 250 32
number of columns in an index 16 10
bytes per row 8060 2000
number of tables in a query 256 32
columns per SELECT statement 4096 255
nested subqueries 32 50
number of enforced relationships 253 32
(obtained from http://www.allinterview.com/showanswers/63402.html[^])

it is better to go with SQL SERVER.

However, cost is also a factor!

But if your requirement is not so high, you can go with SQL SERVER EXPRESS EDITION. which is free

But the license is limited.(hope 10 users)

Hope this answers your question.
Smile | :)

Niladri Biswas

Isaac Gordon30-Jun-09 0:33

Isaac Gordon

30-Jun-09 0:33

thank sir

Hidden tables

mistryshailesh29-Jun-09 21:52

mistryshailesh

29-Jun-09 21:52

Can v create permanent hidden tables in sql 2000 and how?

Re: Hidden tables

Niladri_Biswas29-Jun-09 23:10

OLEDB network failure recovery problem.

29-Jun-09 23:10

Hi,
Though the question is not very clear to me but if you mean that you want to give preference of View access over table access to the users,then the answer is

DENY SELECT ON TABLE TO USER

GRANT SELECT ON VIEW TO USER

If you want to hide the system tables in SQL SERVER 2000, follow this steps:

1) Choose a registered server

2) Right click on that and select Edit SQL Server Registration properties

3)Uncheck

Show system databases and system objects

Hope this helps
Smile | :)

Niladri Biswas

CherezZaboro29-Jun-09 6:57

CherezZaboro

29-Jun-09 6:57

Hi everybody!
I have my own kind of server that for each client keeps an oledb connection opened. It also uses a thread pool for serving the clients. Now let's say the connection between my server and SQL server breaks for whatever reason. While it's down if my server attempts to do anything with the database it obviously fails. Then network issue gets solved.
If my server attempts to continue manipulate the same oledb connection - it still fails. If I reconnect it fails anyway as long as there are those other olddb connections for those other clients of mine. Only when all oledb connections are closed I can start connecting successfully to the SQL server. It seems that in multithreaded/same exe session environment all oledb connections must be closed and reopened in order to recover. Is that true? Or is there a better way to force oledb to revive connections?
Please help!
Thanks a lot in advance.

EncryptbyKey VARCHAR(50)

Saamir29-Jun-09 6:42

Saamir

29-Jun-09 6:42

Hi guys,
I am having a weird issue, I am trying the following and for some reason varchar(50) value is set to null when I encrypt or decrypt it. If I change it from varchar to nvarchar it works. Our table in the database is already set to varchar to store the encrypted information and can't change it due to licensing issues. Please help, what am I doing wrong?

OPEN SYMMETRIC KEY DSBEncryptionKey DECRYPTION BY CERTIFICATE EncryptDsb
DECLARE @encryptedNum varchar(50) -- varchar doesn't work here, nvarchar works

SELECT @encryptedNum=EncryptByKey(Key_GUID('DSBEncryptionKey '), '111000614');
select CONVERT(varchar(50),decryptbykeyautocert(cert_id('EncryptDsb'), NULL, @encryptedNum))

FYI, my symetric key uses AES_256 algorithm.

Please help!

Sameer

My Sql Query Help

Hulicat29-Jun-09 6:26

29-Jun-09 6:26

I have a query that is joining tables and calculating the SUM.

It seems to be working, however; the "SUM" is double what it should be.

Any ideas:

Select, spendaccount, spend.fromcode, spend.keyword,
spend.kwtype, SUM (Rev.trans)
Trans, SUM (rev.newaccounts)
newaccounts, SUM (rev.gross) Gross, SUM (rev.refunds)
Refunds, SUM (rev.net) Net, spend.clicks, spend.spend

From spend, Rev wherespend.fromcode=
rev.fromcode

group by spend.fromcode

Regards,
Hulicat

Niladri_Biswas29-Jun-09 6:54

29-Jun-09 6:54

Hi,
Which SUM you are talking about?

SUM (Rev.trans)?

SUM (rev.newaccounts)?

SUM (rev.gross)?
SUM (rev.refunds)?

SUM (rev.net) Net?

Sniff | :^)

Niladri Biswas

Hulicat29-Jun-09 11:23

Re: My Sql Query Help [modified]

29-Jun-09 11:23

Hello and thanks for the reply; I am not sure on the approach but here is what I am trying to do:

Spend table-> spend (SUM'd), clicks, kytype account
Rev table-> transactions, newaccounts, gross, refunds, net (SUM'd),

From spend, Rev wherespend.fromcode=
rev.fromcode

----all in one output--------

Regards,
Hulicat

Niladri_Biswas29-Jun-09 19:33

29-Jun-09 19:33

I am assuming certain things.

I have created both the tables and has inserted certain values. Please look into it whether you also have the same values or not!

tblSpend

COLUMN NAME			DATATYPE	
-----------			-------
SPENDACCOUNT			int	
FROMCODE			int	
KEYWORD				varchar(50)	
KWTYPE				int	
CLICKS				int	
SPEND				int

Values in tblSpend are

101	1	KeyWord1	1	1	10
102	2	KeyWord2	1	10	11
103	3	KeyWord3	2	21	15
104	4	KeyWord4	3	22	16
105	5	KeyWord5	5	11	11
106	6	KeyWord6	2	33	12
107	7	KeyWord7	1	111	11
108	8	KeyWord8	2	22	2
109	9	KeyWord9	1	1	1
110	10	KeyWord10	1	2	3

tblRev

COLUMN NAME			DATATYPE	
-----------			-------
NEWACCOUNTS			int
GROSS				int
REFUNDS				int
NET				int
TRANS				int
FROMCODE			int

Values in tblRev are

1	10	20	10	10	1
2	20	30	10	20	2
3	30	40	15	25	3
4	40	50	45	5	4
5	50	100	50	50	5
6	60	400	200	200	6
7	70	100	100	0	7
8	70	100	50	50	7
6	111	300	120	180	6

The Query is

SELECT 

		S.SPENDACCOUNT, S.KEYWORD, 
		S.KWTYPE,S.CLICKS,S.SPEND,

		D.FROMCODE,D.TRANS,D.NEWACCOUNTS,
		D.GROSS,D.REFUNDS,D.NET

	FROM (

				SELECT 
						T.FROMCODE,
						SUM (R.TRANS) TRANS,
						SUM (R.NEWACCOUNTS) NEWACCOUNTS,	
						SUM (R.GROSS) GROSS,	
						SUM (R.REFUNDS) REFUNDS,	
						SUM (R.NET) NET

				FROM TBLSPEND T, TBLREV R 
				WHERE 
						T.FROMCODE= R.FROMCODE

				GROUP BY T.FROMCODE 

			) D, TBLSPEND S

WHERE  S.FROMCODE   = D.FROMCODE

OUTPUT:

101	KeyWord1	1	1	10	1	1	10	20	10	10
102	KeyWord2	1	10	11	2	2	20	30	10	20
103	KeyWord3	2	21	15	3	3	30	40	15	25
104	KeyWord4	3	22	16	4	4	40	50	45	5
105	KeyWord5	5	11	11	5	5	50	100	50	50
106	KeyWord6	2	33	12	6	12	171	700	320	380
107	KeyWord7	1	111	11	7	15	140	200	150	50

Hope this helps
Smile | :)

Niladri Biswas

modified on Tuesday, June 30, 2009 4:26 AM

Hulicat1-Jul-09 7:06

1-Jul-09 7:06

Hello Niladri_Biswas, thanks for your help.

Actually I know what the problem is I am just can't figure out the query.

Example of data a problem.

Fromcode Spend.spend rev.gross
ABC 10 70
ABC 20 50
ABC 10 20

Results look like the following:
rev.gross sum'd + spend.spend

Fromcode spend.spend rev.gross
abc 140 (expected 90)
abc 160 (expected 100)
abc etc

I want the total for each column based on the fromcode but not joined sums if that makes sense.

Thanks and Regards

Regards,
Hulicat

smcnulty200029-Jun-09 21:19

smcnulty2000

29-Jun-09 21:19

If you want to solve a problem like this; sometimes you can figure out what is going on by running a similar query to the one you do have but instead of using SUM use COUNT. This is just a debugging method, not the solution.

I suspect if you did that with your original query all of the lines that you saw double would come out with a count of 2 or higher.

Maybe you should try something more like this for your query:

select 
spend.account,
spend.fromcode,
spend.keyword,
spend.kwtype, 
(select sum(rev.trans) from rev where rev.fromcode=spend.fromcode) trans, 
(select sum(rev.newaccounts) from rev where rev.fromcode=spend.fromcode) newaccounts, 
(select sum(rev.gross) from rev where rev.fromcode=spend.fromcode) gross,
(select sum(rev.refunds) from rev where rev.fromcode=spend.fromcode) refunds,
(select sum(rev.net) from rev where rev.fromcode=spend.fromcode) net, 
spend.clicks, 
spend.spend
from spend 
order by spend.fromcode

I don't have mysql running so I can't test it myself.

_____________________________
When life hands you marmots, make marmalade.

Hulicat30-Jun-09 3:59

30-Jun-09 3:59

Thanks "smcnulty2000" and "Niladri_Biswas" your help and input was greatly appreciated.

Regards

Regards,
Hulicat

Hulicat30-Jun-09 7:30

How to Optimize DbTransactions (C#) with SQLServer ?

30-Jun-09 7:30

I am still trying to get it to work, this one didnot return any results

SELECT S.SPENDACCOUNT, S.KEYWORD, S.KWTYPE,S.CLICKS,S.SPEND, D.FROMCODE,D.TRANS,D.NEWACCOUNTS, D.GROSS,D.REFUNDS,D.NET FROM ( SELECT T.FROMCODE, SUM (R.TRANS) TRANS, SUM (R.NEWACCOUNTS) NEWACCOUNTS, SUM (R.GROSS) GROSS, SUM (R.REFUNDS) REFUNDS, SUM (R.NET) NET FROM TBLSPEND T, TBLREV R WHERE T.FROMCODE= R.FROMCODE GROUP BY T.FROMCODE ) D, TBLSPEND WHERE S.FROMCODE = D.FROMCODE

This one "hung" the database:

Thanks...for the help I think I am close

select spend.account,spend.fromcode,spend.keyword,spend.kwtype, (select sum(rev.trans) from rev where rev.fromcode=spend.fromcode) trans, (select sum(rev.newaccounts) from rev where rev.fromcode=spend.fromcode) newaccounts, (select sum(rev.gross) from rev where rev.fromcode=spend.fromcode) gross,(select sum(rev.refunds) from rev where rev.fromcode=spend.fromcode) refunds,(select sum(rev.net) from rev where rev.fromcode=spend.fromcode) net, spend.clicks, spend.spendfrom spend order by spend.fromcode

Regards,
Hulicat

kadaoui el mehdi29-Jun-09 4:55

kadaoui el mehdi

29-Jun-09 4:55

Hello,
I have to Update a Csv file into Sql Server DataBase, the file contains 200 000 records.
So how can i do to optimize the DbTransactions ?! because 200 000 is too big and i want not overload the Database Server.

Solution1: create a global DbTransaction with IsolationLevel=RaadUncommited
Solution2: create a DbTransaction foreach Record with IsolationLevel=RaadUncommited
Solution3: create a DbTransaction for a group of Records with IsolationLevel=RaadUncommited

Already Thanks.

Re: How to Optimize DbTransactions (C#) with SQLServer ?

Enver Maroshi29-Jun-09 7:10

Enver Maroshi

29-Jun-09 7:10

What do you mean by "update" csv into database? You update records in place, or insert new ones?

Look at BULK INSERT at MSDN. It is very fast, you can insert all in temp table and then update records, if that is your goal.

Re: How to Optimize DbTransactions (C#) with SQLServer ?

kadaoui el mehdi29-Jun-09 7:32

kadaoui el mehdi

29-Jun-09 7:32

thank you for you answer.
I update a record.
I know the BULK INSERT but the CSV file have to be in the same machine of Sql Server Data base!

Re: How to Optimize DbTransactions (C#) with SQLServer ?

Niladri_Biswas30-Jun-09 5:50