Re: [h2] corrupt database 1.3.168

2013-10-01 Thread Dieter Cailliau
I assume that that fact (that the h2 version with which i created the schema = 1.3.161) is not relevant for the problem experienced in 1.3.168 which is still a problem in 1.3.168? It is true that i have created a schema once using an older h2, and kept that db file as a file to start of in my

Re: [h2] unable to write: the database is read only after system crash

2013-10-01 Thread Thomas Mueller
Hi, Well, do you use two-phase commit? Could you check if there are any in-doubt transactions, by running select * from information_schema.in_doubt? If yes, those transactions need to be either rolled back or committed first. If no, then I wonder whether you explicitly open the database in

Re: [h2] corrupt database 1.3.168

2013-10-01 Thread Thomas Mueller
Hi, Well, the problem is, if the database was created with an old version, then the corruption might have been there already when you switched to a new version. Corruption problems are very hard to analyze, because there is no log file that tells us which change in the database file was created