Database corrupted during conversion?

[willyvds]

There a tick before this thread, don't know if that means that the matter is solved? Well, sorry: it isn't.

I've installed 1244 on our 4 PC's. I start MM on the first, and make it scan the library. Then I start MM on the second: several errors messages, and then MM crashes.

Some automated messages have gone to support. The debug version is installed, so I can provide more info if required.

We consider this a very serious problem, in fact it means we cannot both use MM simultaneously.

Regards, Willy

[Teknojnky]

I think sqlite is just not very well suited for simultanious multi-user environment.

http://www.sqlite.org/whentouse.html

Client/Server Applications


If you have many client programs accessing a common database over a network, you should consider using a client/server database engine instead of SQLite. SQLite will work over a network filesystem, but because of the latency associated with most network filesystems, performance will not be great. Also, the file locking logic of many network filesystems implementation contains bugs (on both Unix and Windows). If file locking does not work like it should, it might be possible for two or more client programs to modify the same part of the same database at the same time, resulting in database corruption. Because this problem results from bugs in the underlying filesystem implementation, there is nothing SQLite can do to prevent it.

A good rule of thumb is that you should avoid using SQLite in situations where the same database will be accessed simultaneously from many computers over a network filesystem.

What is hosting your database, a windows computer? a nas? a linux box?

If it is a nas, you may be able disable some peformance options regarding write caches/locking files.

From what I understand, the root problem is that whatever is hosting your database is not keeping up with file locking transactions from multiple systems.

Why this worked for you on 3.07 and not 3.1 I can not explain.

Worse case scenario, you re-install 3.07 over the existing 3.1 on all your machines and restore the last usable 3.07 database you have.

Maybe you should check all your mediamonkey.ini files have included:

Code: Select all

[System]
SQLiteSafety=2

[willyvds]

Thanks for the respons.

Indeed, SQLiteSafety = 2. The hosting server is a windows PC (XP).

We did have this problem with 3.0 initially, but after some time the problem "faded out". It occured, but so incidentally that we could accept it.
I don't understand why, though.

We're really very pleased with MM, but with this issue we can't use it as we want. So suggestions are welcome!

Thanks for looking at this!

Regards, Willy

[Teknojnky]

One other suggestion, instead of having multiple computers doing auto-scan, have only 1 do it. That way you greatly reduce the conflicts of multiple updates at the same time.

edit:

regarding 3.0, perhaps the problem re-manifested itself due to the increased database size.

From previous threads, the db increase is primarily due to full text search index added to db.

[rusty]

Also, note that build 3.1.0.1244 includes several changes to improve network performance and error handling.

-Rusty

[willyvds]

I think I get the point: this is not a bug/error/shortcoming in MM, but a result of the combination of underlying database/hardware etc.
So: in fact not a topic for the beta forum. Correct me if I'm wrong!

I'll post a thread in the general forum for further suggestions how to best avoid the problems.

Thanks for the clarification,
Willy

[rusty]

willydvds,

My point was actually to try the new build 1244 since there are changes that are likely to mitigate the problem since there were some network-related problems that were exacerbated in 3.1 (as compared to 3.0).

-Rusty

[willyvds]

Rusty,
yes, I waited until 1244 was available and then tried again immediately. It did not really make a difference.

Note that I am doing this as a test, to try the limits and see if there is an improvement. In normal use, we don't start both MM's with auto-scan on!

But it must be said: when one of us wants to do tagging/synchronising/ripping, we must ask the other to stop MM. Still so with 1244.
And we would like to find a solution for that!

Regards,
Willy