|
I'd recommend to turn off the Auto Close:
ALTER DATABASE database_name
SET AUTO_CLOSE OFF;
GO
Speaking about errors, try to restart the server and run DBCC CHECKDB again. Also, take a look at Windows Event Logs for errors.
BTW, do you have enough disk space?
Regards,
Andrius Leonavicius
|
|
|
|
|
The Log For the Database is Not Available[^]
And please, do not post your question all over the place...
I'm not questioning your powers of observation; I'm merely remarking upon the paradox of asking a masked man who he is. (V)
|
|
|
|
|
hi , i have two table first table name is person and second table name is payment
what i'd like to do is that , i want to select those persons who haven't paid they'r bill for two month
select person.* from (person INNER JOIN
payment ON person.national_number = payment.national_number)
where (payment.payment_date > 'two month ago from now')
but this w'll fetch those persons who paid their bill for the two month ( i don't want them) . but i need those who didn't pay they'r bill for the two month
i don't know if you understand it or no .
|
|
|
|
|
Surely
where (payment.payment_date < 'two month ago from now')
?
|
|
|
|
|
Assuming that one person can have multiple payments, an INNER JOIN will return a record for every payment over two months old, regardless of whether the person has paid within the last two months.
What you need to do is find the person records which don't have a payment record within the last two months. To do that, you can use a Not Exists filter:
DECLARE @Today date = GetUtcDate();
DECLARE @CutoffDate date = DateAdd(month, -2, @Today);
SELECT
*
FROM
person
WHERE
Not Exists
(
SELECT 1
FROM payment
WHERE payment.national_number = person.national_number
And payment.payment_date > @CutoffDate
)
;
"These people looked deep within my soul and assigned me a number based on the order in which I joined."
- Homer
|
|
|
|
|
Won't that have to look up each national_number in payment separately? It always looks like it does.  I don't use EXIST.
You'll never get very far if all you do is follow instructions.
|
|
|
|
|
No, the SQL optimizer is pretty good at generating a sensible query plan. 
"These people looked deep within my soul and assigned me a number based on the order in which I joined."
- Homer
|
|
|
|
|
So..
I would never show up in your list if I paid a dollar each month?
Bastard Programmer from Hell 
If you can't read my code, try converting it here[^]
|
|
|
|
|
Eddy Vluggen wrote: Bastard Programmer from Hell
Oh so appropriate.
Should be renamed The Edge Case from Hell
Never underestimate the power of human stupidity
RAH
|
|
|
|
|
SELECT person.*
FROM person
LEFT OUTER JOIN payment
ON person.national_number = payment.national_number
AND payment_date > 'two month ago from now'
WHERE payment_date IS NULL
You'll never get very far if all you do is follow instructions.
|
|
|
|
|
Hi,
I have employee_number field which has employee_numbers like this:
12
434
ABC345
QWR-567
I want to know how can I just select the numeric values to perform a MAX on it?
so ABC345 and QWR-567 will not be considered in the SELECT
Technology News @ www.JassimRahma.com
|
|
|
|
|
You could always try a regex to parse out only the numbers.
|
|
|
|
|
|
Some guidance on the database you are using would be useful.
Take a look at the ISNUMERIC keyword in BOL.
Never underestimate the power of human stupidity
RAH
|
|
|
|
|
WHERE id NOT LIKE '%[a-z]%' will eliminate any with alphabetic characters, that's a start, work from there.
You'll never get very far if all you do is follow instructions.
|
|
|
|
|
I tried this:
SELECT employee_number FROM employees WHERE employee_number REGEXP ('[0-9]');
it's working excellent and I getting these numbers
7, 8, 9, 11, 1.222, 12, 13, 9, 15, 9, 9, 9
BUT....
when I run:
SELECT MAX(employee_number) FROM employees WHERE employee_number REGEXP ('[0-9]');
I get 9 where i should get 15.
why is that?
Technology News @ www.JassimRahma.com
|
|
|
|
|
Because they're strings; you'll need to convert to numeric fo rthe comparisons to work the way you want.
You'll never get very far if all you do is follow instructions.
|
|
|
|
|
I have been searching all day to find a better answer to my question but those I found was useful
but do not critically analyse my question.
I am building a chat room using php, mysql, jquery and ajax.
The target group members are 3000 people who will frequently
chat every day, so I am expecting like one million messages a day.
The chat room is open to all the members, that means
every member has the same priviledge to send and view
sent messages.
In this case, every member has the permission to delete messages whether sent by him or different members,
however deleted messages should not affect other users. I wouldnt also keep messages for more than two days,
meaning every sent message should be deleted after two days.
Below are the sample tables that represents the logics above
users table
| u_id |
| name |
messages table
| msgID |
| msgBODY |
| msgTime |
| senderID--FK |
| deleted_by_sender|
recipient table
| recID |
| msgID--FK |
| recipientID ---FK |
| deleted_by_recipient|
Now, if I am to implement the schema above, it means that every single sent message has to do 3000 inserts into the
recipient table and one insert into the messages table. This also means that if there are 50 sent messages within 1 minute there would be more
inserts within the one minute. At the same time 3000 people are viewing the messages. Awwsh! more work load on the database within that minute.
hmm!
Please is there a way to handle this, scalability and performance wise?
Any idea is appreciated. Thanks.
|
|
|
|
|
Delete the recipient-table, and add a table like below;
DeletedMessagesPerUser
UserId
MessageId
Add a record to that table when the user "deletes" a message. When fetching all messages for a particular user, fetch all messages where the messageId isn't named in the DeletedMessagesPerUser table.
Bastard Programmer from Hell
If you can't read my code, try converting it here[^]
|
|
|
|
|
Thats a good idea. Please if i want to track unread messages per a user, it means i would have to create another table for the unread messages,right? or there is a better way to deal with it
|
|
|
|
|
Member 10949986 wrote: it means i would have to create another table for the unread messages,right?
Yup - a table with both a reference to the user, and a reference to the messages. Any message-id that's not in the table for that user, would be considered 'unread'.
Bastard Programmer from Hell
If you can't read my code, try converting it here[^]
|
|
|
|
|
Don't you think this would give my queries more load using four tables and doing many JOINs. Cant we have a simple way to solve this, which will increase speed. That even means, at every insert into the message table, there should be the same insert into the unread table, when all members are online
|
|
|
|
|
Databases are optimized to do that stuff
Look into partitioning your table if you need more speed. If that's not enough, then you'd more be thinking about client-side processing. Do you need to "know" on the server whether a user read one of the messages?
If not, then you might be helped quickly with a XMPP-server. The local client could then keep track of the messages it receives, and which ones are read.
Bastard Programmer from Hell
If you can't read my code, try converting it here[^]
|
|
|
|
|
Eddy Vluggen wrote: Do you need to "know" on the server whether a user read one of the messages?
Yes please.
|
|
|
|
|
In that case you can't process that information on the client, and the server will have to.
Bastard Programmer from Hell
If you can't read my code, try converting it here[^]
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
