|
|
Hi all ,
I had developed a project named as Online Examination in Visual Studio 2005 Beta version. Backend for my project is Sqlserver 2000.Which is web based application. So i created a connection object and stored in the Application level.I conducted the exam in a server which have the 100 clients , the server is high end server (IBM server).My examination is divided into number of slots, per day 5 slots.At the first day , after two slots , while in third slot my application had blocked , no button was working. After a R&D i found that a deadlock is occuring in MS Sql server. Can any one tell me the actual problem for this and how to avoid this situation.I am new to Dotnet, this is my first project.
Regards
Srinivasa Rao
|
|
|
|
|
srinuivasaRaoS wrote: Can any one tell me the actual problem for this
No, because you have not provided enough information. (e.g. What code was running on the SQL Server at the time of the deadlock)
srinuivasaRaoS wrote: how to avoid this situation
Typically deadlocks occur because two or more things request the same resource while holding on to other resources that the other wants. For example
Process 1 Process 2
Requests Lock on Table A Requests Lock on Table B
Gets Lock on Table A Gets Lock on Table B
Requests Lock on Table B Requests Lock on Table A
Blocks until Table B released Blocks until Table A is released
Deadlock detection kicks in and one of the processed become the deadlock victim
Deadlock victim:
Lock released on Table B and Transaction Rollsback
Gets Lock on Table B
Completes request.
The solution is to try and ensure that you aquire resources in the same order in each query (easier said than done). So, you can always detect the deadlock in your exception handler (catch block) and re-attempt the query. The SqlException will return a Number of 1205 for a deadlock.
As you are using VS2005 you will have access to the new asyncronous access to the database. You could use this to ensure the application is still responsive while it waits for an answer from the database.
"On two occasions, I have been asked [by members of Parliament], 'Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?' I am not able to rightly apprehend the kind of confusion of ideas that could provoke such a question."
--Charles Babbage (1791-1871)
My: Website | Blog
|
|
|
|
|
Hi Colin Angus Mackay,
Thanks For replying to me,Actually we conducted an online examination which developed in VS2005 Beta version, backend is Sqlserver-2000.I have one server and around 100 clients are connected to the server.My database is also presented in the same server.I am updating the one table for all the clients, which is a status table.And one table for login details.I think this information is enough , if u need still more information , u tell on wht u need , i will tell u the information immedaiately,
Thanks & Regards
Srinivasa Rao
|
|
|
|
|
Then what is happening is that in one transaction you are using the status table and the login table and in some other transaction that is running at the same time you are using the login table and the status table.
I suggest that when you need to use both tables in a single transaction that you use ether the status table and the login table always, or use the login table and the status table always.
"On two occasions, I have been asked [by members of Parliament], 'Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?' I am not able to rightly apprehend the kind of confusion of ideas that could provoke such a question."
--Charles Babbage (1791-1871)
My: Website | Blog
|
|
|
|
|
Thank You colin Angus Mackay,
I have one more doubt , instead of using database table , shell i use xml files for updating the each user status, in the sense one xml file will be created per one user.After that i will write another code which will update the status of each user after completion of exam, is it right way ?.If i will do in this way whether i face any problem.Please Let me know
Thanks & Regards
Srinivasa Rao S
|
|
|
|
|
srinuivasaRaoS wrote: shell i use xml files for updating the each user status, in the sense one xml file will be created per one user.After that i will write another code which will update the status of each user after completion of exam, is it right way ?.If i will do in this way whether i face any problem.Please Let me know
You will just replace row, page and table locking issues for file locking issues. You would be better to figure out what is causing the deadlocks and fix that.
"On two occasions, I have been asked [by members of Parliament], 'Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?' I am not able to rightly apprehend the kind of confusion of ideas that could provoke such a question."
--Charles Babbage (1791-1871)
My: Website | Blog
|
|
|
|
|
My sql storedprocedure contain one SELECT * statement and one RETURN VALUE
Now i want bind the result set of the select statement to datagrid control and return value to the text property of the label control.
Is it possible to bind the data like this?
If possible... HOW??
|
|
|
|
|
Read my article on just such a subject
You need to specific the a SQLparameter as out on your ASP.NET and return records to dataset / datareader.
This is certainly possible.
Look where you want to go not where you don't want to crash.
Bikers Bible
|
|
|
|
|
Hi,
If i am writing simple select query like
Declare @parentInventoryId int
SELECT InventoryID
FROM Inventory_Profile
WHERE
ParentInventoryID=@parentInventoryId
In query execution plan i am getting index scan, but i need index seek
can anybody tell how to change the query to get index seek in my query execution plan
|
|
|
|
|
On what column do you get an indexscan ?
did you make indexes ? on what columns ?
did you make a covering index ?
by adding the requested data and the filtering data in one index, sql does not have to go to the table itself, just the index is enough
how many rows of data are there in this table ?
SQL server ignores indexes when there are few rows of data
My second computer is your linux box.
|
|
|
|
|
I am currently looking for links to papers (articles) on Microsoft Access team development for a research paper. If anyone can direct me to good links or even give some search parameter strings, I would appreciate it. Mean while I will be continuing my search.
Thanks,
INTP
"Program testing can be used to show the presence of bugs, but never to show their absence."Edsger Dijkstra
|
|
|
|
|
John,
What kind of information are you looking for in Microsoft Access team development?
Paul
|
|
|
|
|
Paul,
Thanks for asking, I am afraid that I failed to check back, before writing my paper, after gathering as many relevant articles on team development as I could find via google. As for your question I was looking for any information related to team development, specifically related to Microsoft Access. Even articles related to personal experience (or blogs) in such an environment would have been helpful. I opted to use a more generalized approach, which I believe will meet the requirements. It is not Access specific, but all in all it looks pretty good to me.
INTP
"Program testing can be used to show the presence of bugs, but never to show their absence."Edsger Dijkstra
|
|
|
|
|
Hope your paper does well 
John R. Shaw wrote: personal experience (or blogs)
One of my clients uses MS Access for their sales and marketing lead generation system. From my consulting experience with them, the Access development can be either really nice or really a pain. I personally have a good amount of mixed feelings about Access development. It is nice for putting together a quick rapid prototype, but now with their user load and data load, trying to convince them to move to SQL Server.
Paul
|
|
|
|
|
I am running Windows XP Pro, SQL Server 2005 Express Edition, Visual Studio 2005 Pro.
I need to export my testdb.mdf to Microsoft Excel file. To do this, my friend suggested exporting to flat file first before importing it into Microsoft Excel.
Anyone can teach me the right way to configure / do this? Please help. Urgent. Thank you.
|
|
|
|
|
Hello everybody.I am trying to make an application that uses an ODBC driver to connect to a database of a few related tables.
My questions:
a)Can I embed the odbc drivet to the installation programm so that the user wont have to connect to the database?
b)If I create the related tables in the DBMS but without any records(only the pattern of the database) will the user be able to add and delete his/her own through the connected application?
c)Will I be able to create shophisticated queries only by using the MFC Classes(CRecordset) to manipulate the data?
d)Are there any alternatives except the ODBC and DAO?
-- modified at 8:16 Saturday 20th May, 2006
|
|
|
|
|
Hi,
I have some text which has a lot of newlines. I checked the amount of text in an editor. It showed about 2000 characters. When I tried to save this text in an varchar field of 7000 width an error was generated saying about character limit.
Why is this happening? Is it due to the newline characters.
Thanks
-
बुरा जो देखण मै चला, बुरा न िमलया कोय,
जो मन खोजा आपणा तो मुझसे बुरा न कोय।
Translation
|
|
|
|
|
Dinuj Nath wrote: Why is this happening? Is it due to the newline characters.
Each newline character will count as one character, did you include this count in your original 2000.
You may be better off with a ntext column rather than relying on a varchar
Michael
CP Blog [^] Development Blog [^]
|
|
|
|
|
Thanks for the reply. The count is using the count characters util in scite editor and by the way there won't be more than 20 newline chars. I have been forced to use text and it is eating up space. I just can't understand why I just can't store some thousand simple chars in varchar(7000).
|
|
|
|
|
What version of SQL Server? If SQL 2000, is the row size within the page limit?
In MS SQL Server versions prior to 2005, one row's data is limited to fit in one page or an 8kb block. (But the definition of a table can show a total of more than 8kb per row if you have one or more variable length columns.)
|
|
|
|
|
Thanks a lot. Its Sql Server 2000.
|
|
|
|
|
Hi,
Sorry for my ignorance but could someone give me a hand with this SQL query?
I have two tables in an access database. The first hold a list of titles and the second table holds a list of items belonging to these titles.
I need an SQL statement that will retrieve the titles from table a, and in the same row give me the number of items belonging to that title.
EG:
Table A:
Title 1
Title 2
Title 3
etc
Table b:
item: belongsto:
booksworms Title1
bread Title2
Mice Title3
Cheese Title2
Doorwedges Title2
So my statement needs to return the data as follows, although it needs to be based on the records encountered in table a, rather than simply counting/grouping the records in table b.
Title: Count:
Title1 1
Title2 3
Title3 1
etc
Also if anyone can recommend some useful tutorials id be really appreciative.
Thanks!
|
|
|
|
|
You could use something like,
Select a.*, (select count(*) from tableB b Where b.belongsTo = a.title) as WHATEVER
from TableA as a
Farhan Noor Qureshi
if (this == this) thow this;
|
|
|
|
|
Obviously you have simplified the problem, because there is no need for TableA at all in this scenario; all of the information is contained in TableB.
From TableB only:
SELECT belongsto, COUNT(belongsto)
FROM TableB
GROUP BY belongsto
Assuming it needs both tables due to fields we are not seeing:
SELECT a.Title, COUNT(b.belongsto)
FROM TableA a
INNER JOIN TableB b ON a.Title = b.belongsto
GROUP BY a.Title
MS Access might not like the formatting I have used, and may need some parenthesis around some of that (it would work in SQL Server just the way it is formatted).
I like this book:
Guru's Guide to Trasact SQL by Ken Henderson[^]
But it is SQL Server specific (Transact SQL).
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
