[peterlaursen]

researched on it. Two conclusions:

1) This http://www.webyog.com/whitepaper/Preparing…__Migration.pdf also applies to 'linked' tables.

2) However, after unhiding the metadata, the ODBC-driver does not find any tables in the database! The ODBC-JETengine chain does not seem to be able to resolve OLE/DDE links.

So you can't neither import or export a linked Access DB. But of course you can IMPORT the data from EXCEL into ACCES, save as a .mdb-file and then both import (with SQLyog Migration Tool) and export (from ACCESS) provided that you have the needed ODBC-driver for it.

But the direct migration of CSV's is easier. There also is a problems with EXCEL (and other spreadsheet programs) that it like to 'format' data. For instance any data that can be interpreted as a DATE, will be!

I wrote up the FAQ accordingly http://www.webyog.com/faq/28_73_en.html

[peterlaursen]

Well this

Quote:

2) Consider using ODBC instead. For instance an Excel data file (.xls -file) can be opened from Microsoft Access as a 'datalink' and you can transfer data using ODBC. Then you even don't have to define the columns in advance if the spreadsheet contains the column names in the upper row. ODBC does that for you. And with the SQLyog ODBC/Migration tool you can adjust the datatypes to your need in the import proces.

.. seems not to be so easy! I can then EXPORT from Access (using MyODBC) not IMPORT to MySQL (Using Access-ODBC). But that won't automate things! I'll have to check up on that!

Didn't the direct text-ODBC-import work with comma-CSV's ?

[peterlaursen]

sweet silence <_<

Code:

Label1:
If NO_NEWS == GOOD_NEWS then
{ please tell
}
else GOTO Label1

😀

[peterlaursen]

Can you use this – much simpler – proposal:

You don't need to to link the .CSV's in Access. Simply export from Excel as comma-seperated (semicolon seperated in some countries! http://www.webyog.com/faq/14_88_en.html )

next put all the CSVs in one folder and adress a DSN to that folder from ODBC configuration. The text-ODBC-driver identifies each supported file as a 'table', and the Migration Toolkit lets you import from the folder. Those files (if any) using same SCHEMA can be imported to the same MySQL table if you choose to append data with the Migration Wizard.

This way of doing things is used regularly by some people importing data written by 'low-level' hardware such as industrial PLC's. A few weeks ago I had contact with a customer importing almost 2 GB's of data (about 200 CSV-files) at intervals this way. 100% automatic.

An no … the ODBC driver won't read a Excel docbook!

[peterlaursen]

And if it works you should install the coresponding client package as well. And the MAX server binary if you need it (must be installed over the standard server – adds BDB support among other pretty rare things – as far as I remember InnoDB is included with the standard server in 4.0.x)).

[peterlaursen]

http://dev.mysql.com/doc/refman/4.1/en/news-4-0-14.html

“The CONCAT_WS() function no longer skips empty strings.”

Try this http://downloads.mysql.com/archives/mysql-…0.13-0.i386.rpm

a generic 4.0.13

[peterlaursen]

SHIT!

But it connects then!

Link to arhcives here: http://downloads.mysql.com/archives.php?p=mysql-4.0

changelogs here: http://dev.mysql.com/doc/refman/4.1/en/news.html

see if you can find that implementation in the changelogs and find the latest release before that! I'll research a little too!

[peterlaursen]

Also I believe there are different .RPMs to test with.

Binary RPM's built with a specific Linux version and generic binaries.

. Also source RPM's if you have gnucc-compiler etc installed with the RH Linux.

On my SuSE 10, sometimes the binary versions compiled with SuSE 9.3 will work. Sometimes not. Sometimes the generic will do. Once (Query Browser) I have had to use the source RPMs – neither the SuSE 9.3 version or the generic RPM would work properly. They compile with a single mouse-click on SuSE YaST installer. Probably it is not difficult with YOU either.

A broad variety of builds from here:

ftp://sunsite.dk/mirrors/mysql/Downloads/MySQL-4.0/

This is source: ftp://sunsite.dk/mirrors/mysql/Downloads/&#8230;.0.26-0.src.rpm

This is generic: ftp://sunsite.dk/mirrors/mysql/Downloads/…0.26-0.i386.rpm

Don't forget to tell what you did when you got it working!

[peterlaursen]

I am confused now

Quote:

I'm just syncing the Linux hosted MySql from the BSD hosted MySql

does that mean that it is working now ??

That would be too good to be true, so I don't think that is what you mean!

With SSH-access only there are not so many clients to try .. unless you are allowed to login with a graphical shell (KDE, GNOME) using SSH (don't know RH in this respect, but on SuSE you can)

A few things that you can try if time a saturday night is no issue:

1:

Install 4.0.26 instead of 4.0.12. 4.0.12 is an old boy! – there has been quite a lot of fixes since then, and I would not recommend it. It can be sync'ed with 4.0.12 on BSD as well (same concat_ws() implementation). There should not be but nevertheless there could be an issue with the client library that SJA is compiled with. hmmmm .. but still it connects to SOURCE that is also a 4.0.12 🙁

2:

Do you have a Windows machine available? You can try the SQLyog Enterprise Trial. It won't let you do the sync but it will let you connect with SSH. Just as a test to see if additional info of some kind shows up

[peterlaursen]

Downgrading from 4.1 to 4.0

The MySQL docs has this to say when downgrading from 4.1 to 4.0:

http://dev.mysql.com/doc/refman/4.1/en/dow…ing-to-4-0.html

Anything here that is relevant to your situation?

[peterlaursen]

What is a 'headless' Linux system ?? Sorry but I never heard that before!

I am sorry if I did not get you right.

Your writings must have been too complcated for my brains :huh:

But you are running 4.0 on both systems now I understand. And you can't connect with SJA for LINUX to TARGET MySQL server on local machine. How can you tell ? Error messages in the sja.log?

Are other clients able to connect from local machine? Can SQLyog connect when installed on WINE for instance. 'MySQL Administrator' ? Local MySQL command line client?

Can the SJA connect from another machine (Windows or Linux) ?

Well, I also don't understand the reason for this either!

But you will have to provide more information.

[peterlaursen]

However …

You can also install 4.1 on the source (remember to run the fix_privilege script) and then sync the data.

[peterlaursen]

This message:

Quote:

“The implementation of concat_ws() in source and target MySQL servers are different. SJA uses the concat_ws() MySQL function to generate checksums. Please make sure that the source and target MySQL servers are similar.”

is exlained here: http://www.webyog.com/faq/10_68_en.html

Quote:

This means that the concatenation of colums is performed on the server side, and the checksum's generation and further calculation is done on the client side. Unfortunately the implementation of the concat_ws() function has changed with various MySQL versions. If you try to sync across versions where this is the case, SJA reports an error stating that concat_ws is used and this might not work with different versions of MySQL – and aborts. We have no plans to implement the concatenation on the client side (because that would completely destroy the efficiency of the tool) and thus the sync tool cannot be used if the MySQL versions are “too much different” in this respect. Version 3.23.x cannot be sync'ed with 4.0.x and higher and 4.x cannot be synced with 5.x.

It is NOT a connection issue. It connects! But the DATA SYNC TOOL can't sync across these versions.

You can export your 4.0 database and import to a higher version.

I think your are using the SJA for LINUX. Right? You may install SQLyog on WINE:

http://www.webyog.com/faq/5_71_en.html if you are not able to connect from a Windows client.

And this issue

Quote:

On MySql 4.1 I had a bit of a problem getting “Could not connect to TARGET MySQL server” which I solved by changing the host from “localhost”

probably is a problem with locating the socket file. The MySQL versions uses different sockets (depending on whether it is a dynamically or statically linked version and if you compiled it yourself there are sources available with different compile options). And it is not a SQLyog issue. You would experience the same thing with 'Administrator' and 'Query Browser' as well. You may try!

[peterlaursen]

FAQ write-up: http://www.webyog.com/faq/5_7_en.html

and WINEHQ write-up: http://appdb.winehq.org/appview.php?appId=1328

[peterlaursen]

I must admit that I too think that this looks pretty strange!

I have a music database as one big TABLE 'mp3_files'.

And VIEWs 'title', 'album', 'genre' and 'artist'.

Now let us look at 'artist'. The .sql-file does all this with it.

Is it not highly inefficient? There may be dozens/hundreds of view in a complicated DB-setup! Those that use VIEWs often do so EXTENSIVELY!

Code:

DROP TABLE IF EXISTS `artist`;

DROP VIEW IF EXISTS `artist`;

CREATE TABLE `artist` ……

drop view if exists `artist`;

drop table if exists `artist`;

CREATE ALGORITHM=UNDEFINED DEFINER=`root`@`localhost` SQL SECURITY DEFINER VIEW `artist` AS select ….

Why is it necessary to create the tables at all? It is a waste of server resources (and of time), I think!

[Author]

Posts