I have a pair of databases, one is a live database and one is for testing a configuration for that live database. Both reside on the same server.
I have three tables, Users (PK UserId, FK MainGroupId) and Groups (PK GroupId) and GroupMembers (PK GroupMemberId, FK GroupId and UserId).
The tables are the same schema on both databases however the test database has a set of special test users. Groups is mostly stable, but sometimes we add groups, and sometimes we change column data in the groups. GroupMembers is the same but in the test database refers to the test users.
I need to be able to update the Groups table from the live to test user programmatically. I want to use a bulk copy operation, but to do so I have to delete the Groups table first, which will cause a constraint violation.
I could copy the table in bulk to a dummy table, and then post process by doing an insert of the new rows, and update on the existing rows. However, my problems is that there are about 30 tables like Groups, and I don't want to encode all the column names into the stored procedure in the UPDATE SET method. I'd also like to be able to do it in bulk.
The DBAs are dubious about granting ALTER TABLE permission to temporarily drop the constraints.
Any other suggestions?
SInce both databases are on the same server, why not use a MERGE statement?
select for export and import. If you do it in the right order it should work correctly.
Related
I'm trying to insert data into my database by using a stored procedure but 3 of my columns are using the int identity type and I cannot insert. It keeps saying cannot do this whilst identity insert is off
When IDENTITY_INSERT is on, it just means that you can put your own data in IDENTITY column. It doesn't disable the FK constraint you have on the table. You can delete the FK constraint, or disable it, and risk having logically inconsistent data in your DB, or you can fix your SP so you won't insert any duplicate values.
Something is amiss. Three columns in a single table of type Identity? I'm having difficulty imagining what they could represent, and I have to wonder where the natural keys are.
In any case, IDENTITY_INSERT isn't something you want to putz with casually. It's an administrative feature to allow ad hoc changes to the data, for example bulk loading the database.
If you do actually know what the identities are (as input to your stored procedure) then the table is misdefined, because it's supposed to be the identity source. If you don't know, or you're willing to let the table generate identity values, then you simply don't mention those columns in your INSERT statement. At most, the generated values would be OUTPUT parameters to your stored procedure.
I know you can go into the design view of a table in SQL Server Management Studios and reorder columns as they appear in the design view, however this isn't possible with SQL Azure as the option is disabled. Is there a way to modify SQL Azure tables so that you can reorder their columns as they appear in the design view?
I have been running a number of database upgrades over the last few months to support new requirements and would like to reorder the way the columns appear in design view so they're easier to read, i.e. so they start with a primary key, followed by foreign keys, then normal columns and end with the added by, modified by fields. Its purely to make the tables more readable as I manage them over time.
Just run a script against the table. Its a bit of pseudocode but you should get the idea.
CREATE TABLE TableWithDesiredOrder(PK,FK1,FK2,COL1,COL2)
INSERT INTO TableWithDesiredOrder(PK,FK1,FK2,COL1,COL2....)
SELECT PK,FK1,FK2,COL1,COL2.... FROM OriginalTable
DROP TABLE OriginalTable
Finally Rename the table
sp_Rename TableWithDesiredOrder, OriginalTable
Just another option: I use SQL Delta to propagate my db changes from dev db up to Azure db. So in this case, I just change the col order locally using SSMS GUI, and SQL Delta will do the createnew>copytonew>dropold for me, along with my other local changes. (In Project Options, I set Preserve Column Order=Yes.)
I experienced the same with Azure SQL Database, basically my view changes with ALTER were not taken when did a SELECT * from the view, or the column headers were mixed with the column values.
In order to fix it I dropped the view and re-created it again. That worked.
We have developed a large data migration from one DB schema to the other. We had built it based on the idea that the destination DB would be empty, however months ago we started putting clients on the new application which means their data is being housed in the new schema (the destination DB).
Now we're in a situation where the primary keys could overlap from the source to the destination DB and we're struggling to come up with a solution. The only solution I can think of is to check if the ID exists in the destination, updated the ID in the source to be 1 more than the greatest ID in the destination, and then migrate the record. This seems really cumbersome to have to do for hundreds of tables. Any ideas?
Sorry I don't know anything about SSIS but the following are a few ways to solve the problem using SQL.
When inserting into the destination tables, do not insert identities. As rows are inserted, capture the newly inserted identities and the old identities in a mapping table, see MERGE + OUTPUT INTO. Use the mapping table to update the tables that haven't been inserted, substituting the old identities with the new identities.
Of course for this to work, insertion into tables has to be done in an order that won't cause foreign key or constraint violations.
If you're not into doing all that, and you can lock users out of tables for short periods of time, DBCC CHECK INDENT could be used to 'reserve' identities. These new identities can then be used to update the old data and then insert with SET IDENTITY_INSERT ON.
I am migrating an MS Access application (which has linked tables to a MSSQL Server) to MySQL.
As a means to overcome some MSAccess table naming problems, I am seeking a solution to add a MySQL table alias that will point to an existing table in the MySQL database. Ideally I would like to create the alias 'dbo_customers' in mysql that would point to the customers table also in mysql.
To be clear I am not wanting to alias a table name inside a query like this:
SELECT * FROM customers AS dbo_customers
But rather I would like to be able issue the following query:
SELECT * FROM dbo_customers
and have it return data from the customers table.
Off the top of my head
CREATE VIEW dbo_customers AS
SELECT * FROM customers
Maybe not the best solution but should work as the view is updatable. Will definitely work for Read Only
You can create a View.
CREATE VIEW dbo_customers AS SELECT * FROM customers;
If that doesn't work for you, you could try creating a shadow-copy of the table, and use Triggers to keep the tables synced.
For example:
CREATE TABLE t1( id serial primary key, field varchar(255) not null );
CREATE TABLE dbo_t1( id serial primary key, field varchar(255) not null );
-- INSERT trigger
CREATE TRIGGER t1_dbo_insert AFTER INSERT ON t1
FOR EACH ROW BEGIN
INSERT INTO dbo_t1 SET field = NEW.field;
-- No need to specify the ID, it should stay in-sync
END
-- UPDATE trigger
CREATE TRIGGER t1_dbo_update AFTER UPDATE ON t1
FOR EACH ROW BEGIN
UPDATE dbo_t1 SET field = NEW.field WHERE id = NEW.id;
END
-- DELETE trigger
CREATE TRIGGER t1_dbo_delete AFTER DELETE ON t1
FOR EACH ROW BEGIN
DELETE FROM dbo_t1 WHERE id = OLD.id;
END
Not exactly an 'alias', and far from perfect. But it is an option if all else fails.
there is a simpler solution for MySQL via MERGE table engine:
imagine we have table named rus_vacancies and need its English equivalent
create table eng_vacancies select * from rus_vacancies;
delete from eng_vacancies;
alter table eng_vacancies ENGINE=MERGE;
alter table eng_vacancies UNION=(rus_vacancies);
now table rus_vacancies equals to table eng_vacancies for any read-write operations
one limitation - original table must have ENGINE=MyISAM (it can be easily done by "alter table rus_vacancies ENGINE=MyISAM")
You could create a view named dbo_customers which is backed by the customers table.
#OMG Ponies ponies said in a comment:
Why not rename the table?
...and it seems the obvious answer to me.
If you create an ODBC linked table for the MySQL table customers it will be called customers and then all you have to do is rename the table to dbo_customers. There is absolutely no need that I can see to create a view in MySQL for this purpose.
That said, I'd hate to have an Access app that was using SQL Server table names when the MySQL tables were not named the same thing -- that's just confusing and will lead to maintenance problems (i.e., it's simpler for the linked tables in the Access front end to have the same names as the MySQL tables, wherever possible). If I were in your position, I'd get a search and replace utility and replace all the SQL Server table names with the MySQL table names throughout the entire Access front end. You'd likely have to do it one table at a time, but in my opinion, the time it takes to do this now is going to be more than made up for in clarity going forward with development of the Access front end.
I always rename my "linked to SQL" tables in Access from
{dbo_NAME} to {NAME}.
The link creates the table name as {dbo_NAME} but access occasionally has problems with the dbo_ prefix.
Aliases would be nice, yet MySQL does NOT have such a feature.
One option that may serve your needs, besides creating a view, is to use the FEDERATED storage engine locally.
CREATE TABLE dbo_customers (
id INT(20) NOT NULL AUTO_INCREMENT,
name VARCHAR(32) NOT NULL DEFAULT '',
PRIMARY KEY (id),
)
ENGINE=FEDERATED
DEFAULT CHARSET=latin1
CONNECTION='mysql://fed_user#localhost:9306/federated/customers';
There are currently some limitations with the FEDERATED storage engine. Here are a couple especially important ones:
FEDERATED tables do not support transactions
FEDERATED tables do not work with the query cache
I'd like to mention a bad solution I explored (and abandoned), which was to use hardlinks on the .frm, .MYD and .MYI files corresponding to my table in /var/lib/mysql/{name_of_my_database}/.
It does, however, NOT work. For InnoDB tables, it simply cannot (even if you hardlink the .idb file) because tables are also referenced in ibdata1.
For MyISAM tables, it kind of works, except it doesn't because in memory, the tables are still distinct and thus do not share cache. So if you write a row to original_table, it won't immediately appear in aliased_table. You would have to flush tables first… which defeats the purpose and even causes data loss (if you insert a row in both the original and the alias before flushing, only one row is kept).
I thought my experiment was worth mentioning as a cautionary tale.
I have written the following trigger in SQL server:
create trigger test_trigger
on invoice -- This is the invoice table
for insert
as
declare #invoiceAmount int -- This is the amount specified in the invoice
declare #custNumber int -- This is the customer's id
--use the 'inserted' keyword to access the values inserted into the invoice table
select #invoiceAmount = Inv_Amt from inserted
select #custNumber = cust_num from inserted
update customer
set amount = #invoiceAmount
where Id = #custNumber
Will this be able to run in MS Access or is the syntax different?
The Access database engine (formerly called Jet) does not have triggers and regardless has no control-of-flow syntax e.g. a PROCEDURE must consist of exactly one SQL statement.
Tell us what you really want to do and there could be an alternative syntax.
For example, you could create a new key using a UNIQUE constraint on invoice, (cust_num, Inv_Amt), a FOREIGN KEY customer (id, amount) to reference the new key, a VIEW that joins the two tables on the FOREIGN KEY columns and exposing all four columns, then INSERT into the VIEW rather than the table 'invoice'; you may want to use privileges to prevent INSERTs to the base table but user level security was removed from the new Access 2007 engine (called ACE).
But, if you don’t mind me saying, I think your trigger doesn't reflect a real life scenario. A column vaguely named 'amount' in table 'customer' to hold the most recent invoice amount? What about when the inserted logical table contains rows for more than one customer? As I say, I think you need to tell us what you are really trying to achieve.
Access doesn't have triggers
Your trigger that you show here will bomb out since it does not take into account multirow updates the moment someone updates more than one row (and don't say it won't happen because it will better to practice some defensive coding)
Triggers fire per batch not per row, please read Multirow Considerations for DML Triggers
join inserted pseudo table and the invoice table instead to update the values...that works for 1 and more than 1 row
They may be coming in Access 2010? http://blogs.msdn.com/access/archive/2009/08/13/access-2010-data-macros-similar-to-triggers.aspx
MS Access doesn't have triggers.
That is, the the Access Jet engine (which creates .mdb files). If Access is connecting to a database server, then it will use whatever triggers are in that database.
I've never come across triggers in Access unless it's dealing with ADP on SQL Server. So your answer is yes, it's the same if you're on SQL Server for the backend, and no if the table is stored in Access.