Devart Forums

I have problems with the batch updates again ... (you fixed this bug in the 10.4.3 version)
But:
From time to time I get EAccess Errors when I send the Execute command to start the batch update to my remote servers.
After that the program will crash with different errors which indicates to some memory allocations problems.

Could you fix this?

Hi Tobias,

Unfortunately, we could not reproduce described error on our environment- our tests do not locate any errors and no one reported this issue before.
In order to investigate described behavior, we need to reproduce it on our side. Please compose a sample which demonstrates this behavior including scripts to create and fill in database objects and send it to us using contact form from our website:
https://www.devart.com/company/contactform.html
As soon as we get this sample and if the reason is in the code of our product, we will try fixing it asap.

Regards,
Viktor

Last time I reported bugs you could not reproduce the error either. And I was the only one who had this issue. But the problem is real... I can´t help you out whith a sample app.

Hi Tobias!

Thank you for your reply.

After your previous request in this topic viewtopic.php?f=7&t=44551 we immediately reproduced the issue.
Kindly note that earlier viewtopic.php?f=7&t=44543 we could not reproduce the behavior based on the data provided in the beginning of the post, so we asked you to provide an example.
After you have provided us with the missing information (LockMode = lmPessimistic) to reproduce the issue in our environment, we immediately fixed it after reproducing.
That is why we ask you to send a sample that stably reproduces the behavior you indicated, since it is the only way to reproduce and fix it.

Looking forward to hearing from you!

Regards,
Viktor

The problems with the batch updates started when you "optimised" the code to speed up the uploads. Maybe you can have a look at this part of the code. It ssems like there are access violations in the array.

Hi Tobias,

Before release of any build, we run our tests. Only if all tests were successful, we release the version. If, after the change, new issues in code are located- after the reproduction, we fix them and as result new bugs are covered by new tests.
As we mentioned before, we cannot reproduce the issue on our environment and as a result ask for a sample because this is the only way to resolve this issue.

Regards,
Vikto

What should I say? I did not change a single line of my code, but after updating from 10.4.1 to 10.4.3 I have this problem...
Ok thanks for your help.

Hi Tobias!

Thank you for your reply!
Please be informed that we tried to change our tests, but there are no errors, so we cannot re-create your situation.
Could you please provide us with the following information, so that we could simulate such an environment on our side: a script for creating a table, a set of actions which began to lead to the error, the exact error message that you receive, a set and size of batches, what batches you are running (insert, update ...) etc.
Please, also specify the version of the server you are using: Ver 10.4.14-MariaDB-1:10.4.14 or another one.
Looking forward to your reply.

Regards,
Viktor

I am now using the latest update (11.0.1) the problem is gone. Magic?

Sorry - the problem still exists - today I got the error again ...

Hi Tobias,

Not good... In this case we need a sampe which reproduces the issue on permanent basis.

Since you are the first person who reported this we need to perform more testing in order to identify the issue- because all previous tests we performed could not reproduce the issue on our side and without a sample from your side we will not be able to locate it. .

Thanks in advance,
Viktor

This error started, when you "speed up" batch inserts. Could you provide an option to use the old (slower) version? Maybe one could use a parameter to switch between the two versions. I get the error every single day - this is very disappointing to me.
Often I get an error : commands out of sync.

Dear Tobias,

Thanks for your reply!
When we pointed out that we had improved the work of batches, we carried out multiple tests in our environment and received a result of improving the work of batches. But at the same time, the version with this improvement included many changes and additions, not only at the MyDAC level, but also at the basic level.
Also, as we have already written many times, you are the only user with a similar problem. And the only solution for solving it is to reproduce it in our environment. As soon as we can reproduce it at home, we will immediately begin to study the causes and correct the error.
Please compose a sample which demonstrates this behavior including scripts to create and fill in database objects and send it to us using contact form from our website:
https://www.devart.com/company/contactform.html

Should you have any questions, do not hesitate to ask!

That is correct - but do you remember that I already found two bugs?

Now I use the dbMonitor and got this result which leads to the described error when I execute the batch insert.
when the error occurs the out value is missing


this is how it looks like, when the update is successful