I'm considering a MySQL to Postgresql migration for my web application, but I'm having a really hard time converting my existing MySQL database to Postgresql.
I tried :
mysldump with --compatible=postgresql
migration wizard from EnterpriseDB
Postgresql Data Wizard from EMS
DBConvert from DMSoft
and NONE of the above programs do a good job converting my database!
I saw some Perl and Python scripts for converting mysql to postgresql, but I can't figure out how to use them....(I installed ActivePerl and don't understand what I'm supposed to do next to run that script!)
I use Auto Increment fields (as a primary key) all the time, and these are just ignored... I understand that Postgresql does auto-increments in another way (with sequences), but it can't be THAT hard for MIGRATION software to implement that, or is it?
Did anybody have better luck converting a MySQL database with auto-increments as primary keys?
I know this is probably not the answer you are looking for, but: I don't believe in "automated" migration tools.
Take your existing SQL Scripts that create your database schema, do a search and replace for the necessary data types (autonumber maps to serial which does all the sequence handling automagically for you), remove all the "engine=" stuff and then run the new script against Postgres.
Dump the old database into flat files and import them into the target.
I have done this several times with sample databases that were intended for MySQL and it really doesn't take that long.
Probably just as long as trying all the different "automated" tools.
Why not use an ETL Tool? you dont have to worry about dumps or stuff like that.
I have migrated to PostgresSQL and MySQL and have had no problems with the auto increment fields.
You just need to know the connection credentials and thats it. I personally use Pentaho ( it's open source ).
Download Pentaho ETL from http://kettle.pentaho.org/
Unzip and run Pentaho (using .bat file spoon.bat)
Create a new Job:
Create DB connection for source data base (PostgreSQL) - using menu: Tools→Wizard→Create DataBase Connection (F3) Create DB connection for destination data base (Mysql) - using technique described above.
Run the Wizard: Tools → Wizard → Copy Tables (Ctrl-F10).
Select source (left dialog panel), and destination (left dialog panel). Click Finish.
The Job will be generated - Run the job.
If you need any help let me know.
Even when you familiar with all "PostgreSQL gotchas", doing every step by hand may take a lot of time, especially when your db is "big".
Try some other scripts/tools.
I know this is an old question but I just ran into the same problem migrating from MySQL to Postgres. After trying several migration tools out the very best one I could find, which will migrate your database structure as cleanly as possible, was Pgloader https://github.com/dimitri/pgloader/ it will take care of changing the Auto Increment to Postgres sequences no problem and it's super fast.
Related
I currently have a database in MySQL, which I'd like to import in MS Access.
Is it possible to do this while keeping all relationships intact (i.e. without exporting to .csv, or by using ODBC)?
I'm a noob in this area so any help is greatly appreciated.
Thanks.
You need to solve two different problems:
Creating an empty MS Access database with a structure that matches the MySQL database structure.
Extracting the data from MySQL and loading it into MS Access.
This is not easy because different SQL databases offer different structural features, different datatypes, and so on. The more complex your use of MySQL is the more likely you'll run into some show-stopper during the conversion (for instance, Access doesn't support triggers at all). Conversely if you're using MySQL as a simple data store you may find the conversion fairly easy.
To get an MS Access database with the same structure as your MySQL database, your best bet is to find a database definition / diagramming tool that offers reverse engineering and supports both MySQL and MS Access. Use it to reverse engineer your MySQL database into a database diagram, then change the underlying database to MS Access and use the tool to generate a database.
Check out Dezign For Databases which (on paper, anyway) offers the features you would need to do this.
To pump the data across, there are any number of tools. This kind of operation is generically referred to as ETL (Extract, Translate, Load).
Do you mean SQL Server? A good starting point might be to check out SQL Server Integration Services (SSIS), which can be used for transferring data around like that.
Google will also be helpful, check out the first result:
http://support.microsoft.com/kb/237980
By the way, you said ".sql" in your question: a .SQL file is a script file, which could do anything from create a database, insert data, drop table, delete data, or given the right permissions, call system procedures and reboot a machine, format a drive, send an email.. Just for ref, .SQL files aren't the storage format used by SQL Server.
While you can script your database's schema into script files via something like SQLyog, you will find that the syntax varies enough from database to database (MySQL to Access, in your case) that you can't directly apply the scripts.
With much effort a conversion script could be created by editing the script (perhaps automated with a program, depending on the resulting script size). I think you would be better served using ODBC to copy the tables (and data) and then extracting and re-applying the relationships from the generated script by hand. Time consuming, but also a one time operation I would hope.
When both systems are the same database, there are tools that can do the comparison and script generation (TOAD for MySQL and RedGate Compare for Microsoft SQL), but they don't do cross database work (at least not the ones I am aware of).
If you create a ODBC DSN, you can use TransferDatabase to import from your MySQL database. You can do it manually with the GET EXTERNAL DATA command (or whatever it is in A2007/A2010) and see how well it works. It won't get all data types exactly right, but you could do some massaging and likely get it closer to what will work best.
Is there some reason you can't just link to the MySQL tables and use them directly? That is, why do you need to import into Access at all?
Access: run query. Just make sure to adapt the SQL code since every RDMS has its own sintaxis (despite SQL being an ANSI standard).
I'm creating a new Asp.Net MVC 3 application. Visual Studio does a lot of the job of create the database and initial layout. Very nice! I will upload that initial files to my server, but I want that it runs using the MySql database on the server.
There's some quick/easy way to do it? I'm not worried about the data, just the structure of the tables, and the connection/configuration changes.
Thank you very much!
You can export any MS-SQL database as a Script (Sql Server manager).
Fix it up to make it compatible.
But you will also need a Membership provider, look around if there exist any for MySql, otherwise you'll have to create one (movie).
There are a number of tools listed in "Migrating from Microsoft SQL Server and Access to MySQL".
Or (assuming that you're using column types that exist on both platforms) you can write a script to convert a schema dump from SqlServer into MySQL (or do the conversion by hand in a text editor). Even better yet, you can write a program program to read the INFORMATION_SCHEMA table from SqlServer and produce the necessary CREATE TABLE... statements in mysql. Lots of options.
I am trying to replicate a Progress database to MySQL 5.1..now , I came across a few softwares and a few suggestions on stackoverflow as well as other websites which necessitate the involvement of a Software like Pro2SQL or other SQL migration tools like MySQL migration tool.But the problem that i am faced with is that I will be using Linux to run the mysql.i am working on linux.Is there a software for linux(I am using bash scripting to query the MYSQL database) or another other means?
Currently , I am using jdbc to connect and retrieve, but mapping the database is hard and may create flaws in the long run due to mapping problems.Also, this proccess will be repeated quiet often..for backup.
Since, MySQL migration tool is a good solution , but it doesnt support linux command prompt, so I have to implement in another better / optimized way..Please suggest what should be done further.Thanks a ton for the support..
If it is just about dumping :
If i get your problem the solution holds in two lines (If you are following SQL standards) :
pg_dump <yourdatabase>
mysql < <yourfile.sql>
With the first line you are dumping your database, many options exists whether you want to dump tables, content, schemas, etc... go to the man page for more details
With the second lines you are just loading them into your mysql.
If it is about mapping :
Take a look a Kettle, it's an Open Source ETL, it works really well on Linux and you can automize task using crontabs.
Hope I could help,
I have a database (mdb file) that I am currently busy with. I would like to know if it is possible to generate MySQL code that would be used to create this database?
There are a couple of tools you can look at to try to do the conversion.
DataPump
Microsoft DTS (Nos Called SQL Server Integration Services)
Other option might be generate MySQL code from Access' DB MetaData you can access from JDBC, ODBC, ADO.NET or any other database access technology with metadata support. For this option you need to generate a piece of code (script). So it will only make sense if your access DataBase has a lot of table with a lot of columns or if you are planning to do this task several times.
Of course, using one of the mentioned tools will be faster if it works.
You can certainly write DDL to create and populate a MySQL database from the work that you've already done on Microsoft Access. Just put it in a text file that you execute using MySQL batch and you're all set.
If you intend to keep going with developing both, you'll want to think about how you'll keep the two in synch.
The .dmp is a dump of a table built in Oracle 10g (Express Edition) and one of the fields is of CLOB type.
I was trying to simply export the table to xml/csv files then import it to the MySql, but the export simply ignored the CLOB field... (I was using sqldeveloper for that).
I noticed this post explaining how to extract the CLOB to text file but it seems to miss the handling of the other fields or at least the primary key fields. can it be adopted to create a csv of the complete table? (I am not familiar with plsql at all)
As the brute force approach, I can use my python interface to simply query for all the records and spool it to a flat file but I'm afraid it will take a LOOOONG time (query for all records replace all native commas with the ascii... )
Thanks guys!
if you can get the mysql server and the oracle server on the same network, you might want to look at the mysql administrator tools, which includes the migration toolkit. you can connect to the oracle server with the migration toolkit and it will automatically create tables and move data for you.
Here is a documentation explaining the migration process: http://www.mysql.com/why-mysql/white-papers/mysql_wp_oracle2mysql.php
and you can use Data Wizard for MySQL . Trial version is fully usable for 30 days.
After about 2 hours of installing and uninstalling the MySql on the same machine (mylaptop) in order to use the migration tool kit as suggested by longneck, I decided to simply implement the dump and here it is for the likes of me that have minimal admin experience and get hard time to make both DBs work together (errors 1130, 1045 and more).
Surprisingly, it is not as slow as I expected: OraDump
Any comments and improvements are welcomed.