[royvarley]

Tried with 4.07 download. SJA reports V4.1

Same problem.

Note that the error message puts single quotes around the text where it thinks there's an error. So the leading quote at 's Quarry and the trailing quote at Sectio' were supplied by SJA error reporting. '15' is ok – it's common practice to quote numerics in a SQL statement.

[royvarley]

SJA 4.1

SQLyog 4.07 download

What should we forget? 🙄

[royvarley]

Sorry – SJA 4.05

In this case, Columns selection and SQL Where not used.

But since you say that should have been fixed at 4.06/07, I guess I'll try that.

[royvarley]

I'm getting similar problem on sync jobs.

When a data item contains ' I get an error in the log file:

Error No. 1064

You have an error in your SQL syntax near 's Quarry', `CostumeDescription`=NULL, `CompanyID`='1', `ProjectID`='15', `Sectio' at line 1

The data in the field was Mr Slate's Quarry

The record containing this field does not update.

[royvarley]

Well, strictly speaking, a key can be one, or a combination of more than one, field in a table. A primary key just needs to be unique and it's valid to have more than one field to achieve that.

In my experience, I just find life a lot easier if a primary key is a generated unique number. Either using an autonumber field or having a special routine to generate a unique number. You can then create indexes on other fields where it makes sense for common queries into the data. Doing this makes using a pk as a foreign key easy. In your case, for example, for a record in another table to refer uniquely to a record in the example table you've given us, requires holding (and managing in the software) four fields.

Then again, I don't know your application or the design trauma you've had to go through to get to where you are at. If it makes sense for you, don't let me sway you into changing things for the sake of it. And that's not why you're here asking for help, is it! 🙄

[royvarley]

Hmmm… looks like a job for the folks at webyog!

However, even though you say the mysql versions are the same, I agree with Peter's comment about the wisdom of using TIMESTAMPS as part of the pk. I'd re-visit that part of the design if possible. Maybe use a simple long integer as the pk and then construct alternate keys using either TIMESTAMP or simple strings based on date/time down to the resolution you need.

[royvarley]

Thanks, Ritesh. I'll send you the details soon. I don't have the original but I'll construct something simple to illustrate. Bear with me – just got snowed under…

[royvarley]

Do you specify any columns or use a SQL Where clause?

[royvarley]

oops – I just realised that I'd said SJA does structure synch – and you said it doesn't – got it! Not till 4.2. OK – I'm with it now. 😀

[royvarley]

Quote:

QUOTE  

SJA.exe runs for all job types including synchronization. So you can run just about any sort of query as well as data and structure synchronisation. The xml defines the type of job.

Currently, SJA only supports Data Synchronization and Notification Services. Notification services allows you to execute any query(s) and send the result over mail.

Yep – glad I got that right. But just to clarify for everyone – Notification services type jobs can run a query without having to send any form of notification. I'm currently using this for some nightly maintenance tasks.

Quote:

QUOTE  

as far as I can tell, it doesn't handle running mysqldump to create the .sql files.

Nope it dose not.

In v4.1, SJA will support dumping of data in .sql files like mysqldump.

Great! Any ideas when 4.1 SJA for windows will be available?

Quote:

QUOTE  

I use the query BACKUP TABLE {list of tables} TO {directory}; The query was parsed ok.

SJA logs all the errors in sja.log found in the installation directory of SJA. What does sja.log say?

It just says: “Job started at Sun Feb 06 15:31:59 2005”. No error messages at all. Hence my remarks about the level of both error reporting and simple logging of job details.

Thanks for that Ritesh. Guess I'll just hang on till SJA 4.1

[royvarley]

Quote:

Thats what SJA is for, and NOTHING ELSE !!!

Well… not quite. Under windows, SJA.exe runs for all job types including synchronization. So you can run just about any sort of query as well as data and structure synchronisation. The xml defines the type of job.

That said, I'm thinking of using the export functionality – but now I hit a limitation of sja – as far as I can tell, it doesn't handle running mysqldump to create the .sql files. Unless, of course, it does… Ritesh?

[Author]

Posts