|
Yes. I have to have a solution that does not require any licensing or cost. The Information has to be updated automatically with some frequency (say every 5 to 15 minutes) so that when all 7 Stand Alone Servers have the same information and new records and record changes in all of there databases which would be identical in nature. It can be accomplished with any free tool or package that works with MySQL 5.1. They all communicate with each other and I granted access for each of them for Replication and set up Relay and Binary Log Files as is required with Replication but none of the servers get updated when new records are entered in any server or changed. If a Backup is done on any Server it takes less than a minute. If it can be Restored on the Next Server in the Change and Vice Versa, it should be OK as long as it can be deployed automatically and it is staggered so there is no Writing Conflicts.
Regards
BC
|
|
|
|
|
I'm sorry, but I don't get it. Consider for example following timeline:
Server A Server B
-------- ---------
insert 1 |
| insert 2
| |
Backup 1 |
| insert 3
| |
| Restore 1
When Backup 1 is restored in Restore 1, inserts 2 and 3 are lost.
|
|
|
|
|
Data Loss is not critical. If Migration is used every 5 minutes they all synchronize with each other fine. Each server is in a stand alone world. The other servers information is not critical as long as it gets there eventually.
Regards
BC
|
|
|
|
|
So I guess you have 7 databases on each machine.
Anyhow, the backup is taken using Backup[^] command and then it's restored using Restore[^] command. You can schedule these using for example Windows Task Scheduler and a command script.
|
|
|
|
|
The backup must be done automatically on Server A and then Restored Immediately Afterwards on Server B until all 7 Servers have the Previous Servers Information. When Server B completes its Restore from Server A it must Start a Backup and complete the Sequence chain until all are updated automatically. There are 10 databases on Each Server that are identical in nature.
Regards
BC
|
|
|
|
|
Only the information and new records from the Previous Server in the Change need to be Restored into the Next Servers Databases.
|
|
|
|
|
Then you have to use --log-bin option and mysqlbinlog [^] utility.
What comes to the other question about the sequence of operations on different server, there's no native support for that in MySQL so you can for example create one centralized script which does all the backups and restore operations on all servers. Or if you want to create a script on each server you must create a mechanism how to synchronize the operations (using for example your master database to store information what is being done and what's next).
|
|
|
|
|
I am using a log-bin file on each of the Servers and Each Server is a Slave of the Previous Server with the Last One in the Chain is the the Sever for the Save Number 1. any Topology will be fine as long as all the data is the same in all the Servers at the end of the day.
|
|
|
|
|
Hi,
I have a stored procedure which firstly inserts a record into a table and then uses a cursor to update all the records in the table (there is a field called sequence, which specifies the order that the records are retrived, that needs to be updating).
The problem I seem to have is that the cursor updates all the records except for the one that has been newly inserted. Although my Insert statement comes before the cursor, it seems as though the record is being inserted after the cursor has finished.
Any ideas anyone?
(I am using SQL Server 2005)
|
|
|
|
|
If you could post the code, it would greately help.
Perhaps you open the cursor before the insert statement or yu have where condition which eliminates the inserted row.
Another thing, I believe it would be much better, in terms of speed, if you would do the updating without cursors. Perhaps the cursor operation could be transformed to a set based update statemnt.
|
|
|
|
|
Thanks for your time.
I understand what you're saying about cursors being a performance hog, but there are going to be at most 30 records in this table so I didn't think using a cursor would have too much of an impact. Am I wrong in thinkign this?
Anyway, here is my procedure:
ALTER PROCEDURE [dbo].[spReportTabINSERT]
-- Add the parameters for the stored procedure here
@prmName as varchar(30),
@prmSequence as int
AS
BEGIN
SET NOCOUNT ON;
INSERT INTO
tblReportTabs
(
[name],
[Sequence]
)
VALUES
(
@prmName,
@prmSequence
)
-- update all sequences
DECLARE @ID uniqueidentifier
DECLARE @Sequence int
DECLARE @NewSequence int
SET @NewSequence = 10
DECLARE crsTabs CURSOR
FOR
SELECT
tblReportTabs.[ID],
tblReportTabs.[Sequence]
FROM
tblReportTabs
ORDER BY
tblReportTabs.[Sequence]
OPEN crsTabs
FETCH NEXT FROM crsTabs INTO @ID, @Sequence
WHILE @@FETCH_STATUS = 0
BEGIN
UPDATE
tblReportTabs
SET
tblReportTabs.[Sequence] = @NewSequence
WHERE
tblReportTabs.[ID] = @ID
SET @NewSequence = @NewSequence + 10
FETCH NEXT FROM crsTabs
END
END
CLOSE crsTabs
DEALLOCATE crsTabs
|
|
|
|
|
The first observation is that in the end of the loop, you should have:
FETCH NEXT FROM crsTabs INTO @ID, @Sequence
instead of
FETCH NEXT FROM crsTabs
|
|
|
|
|
Great, fixed!! School boy error aye?!
Thanks a lot 
|
|
|
|
|
No problem.
Liqz wrote: School boy error aye
It happens to everyone
Another thing, I didn't quite get the logic in your loop. You fetched the sequence but you didn't use it anywhere, was that intentional?
|
|
|
|
|
Again, you're right. I am updating the sequence with @NewSequence, but of course, I do not need to fetch it in order to do this. I will amend it.
Thanks again
|
|
|
|
|
You're welcome
|
|
|
|
|
I must be off my rocker but I can't see why you are using a cursor for this logic.
Need software developed? Offering C# development all over the United States, ERL GLOBAL, Inc is the only call you will have to make.
If you don't ask questions the answers won't stand in your way.
Most of this sig is for Google, not ego.
|
|
|
|
|
Because I need to loop through each record in a table and update a field based on an incrementing integer.
I'm failry new to all this so if you have any other ideas then I'm all ears!
|
|
|
|
|
Well autonumbers are the way to do incrementing integers in Sql Server and sequences in Oracle (look-up either SCOPE_IDENTITY() or sequence NextVal depending on your poison of choice)
However, to get your code to work it seems to me:
1) Begin Transaction
2) Insert
3) UPDATE set sequence = sequence + 10 WHERE sequence > @sequence
4) UPDATE set sequence = sequence + 10 where id = @id
5) COMMIT
CURSORS are evil and should be avoided.
Need software developed? Offering C# development all over the United States, ERL GLOBAL, Inc is the only call you will have to make.
If you don't ask questions the answers won't stand in your way.
Most of this sig is for Google, not ego.
|
|
|
|
|
Thanks for your response. I understand your concern about cursors but unfortunately I don't think your idea is going to work. The reason is because the insert statement is inserting a record with a sequence which isn't a multiple of 10, i.e. 15, 25, 35 etc. This then means that the newly inserted record is in the right place accoring to sequence number. The cursor then gets all the records ordered by sequence, and updates them back to numbers divisible by 10, including the newly inserted record (obviously, the sequence doesn't have to be multiples of 10, it could of easily have been 2, but still).
Again, I realise that cursors are not a good idea but I don't see any other way around it (other than doing it from my VB.NET application with multiple database calls which definately isn't a good idea). Also, there are going to be at most 30 records in this table so I thought that using a cursor wouldn't be too much of a problem.
Like I said, I am fairly new to this so any other suggestions you may have are welcomed.
|
|
|
|
|
I highly recommend you reevaluate your logic in your application and database. I think you will find the answer should come to you. While I do not know what you are doing, I have a pretty good idea.
Need software developed? Offering C# development all over the United States, ERL GLOBAL, Inc is the only call you will have to make.
If you don't ask questions the answers won't stand in your way.
Most of this sig is for Google, not ego.
|
|
|
|
|
Ok Thanks. I have just looked at your example again and now understand a little more where you're coming from and i think that it would acually work for my insert statement by changing one of the lines from
UPDATE set sequence = sequence + 10 where id = @id
to
UPDATE set sequence = sequence + 5 where id = @id
which is great and I will use this, thanks
But I have a similar stored procedure for doing an update on an existing record, that is, updating the sequence of one of the records already in the table. Maybe I will look into my logic further for this too.
Thanks again
|
|
|
|
|
I understood that you're trying to do ordering. For example you have ordinals 1,2,3,4 and 5 then you add a new record which has the ordinal 3 so old ordinals 3,4 and 5 will be replaced with 4,5 and 6 respectively. If that's the case, you could simply:
- update all rows, set ordinal = ordinal + 1 where ordinal >= 3
- insert the new row with ordinal 3
|
|
|
|
|
That is actually a good and very simple idea too, why didn't I think of that!
Will try that when I get to work tomorrow!
Thanks
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
