I assume so. As I wrote before. I had to cancel the concurrent Part of the AbstractES-Test for Swift as it didn't return at all. I think when canceling the test the store became corrupt. So the question is - does our commit protocol help us there? Perhaps have a "transactional" wrapper around the RandomAccessFile that buffers writes until a commit command comes? This will become interesting with multiple threads but I assume a thread local buffer would help there and a synchronized write of all the committed buffers.
Michael Am 01.03.2009 3:37 Uhr, schrieb Niclas Hedhman: > On Sat, Feb 28, 2009 at 6:21 PM, Michael Hunger<[email protected]> wrote: >> MacBook Air with MacOS X 10.5 >> 1.6.0_07-b06-153 > > Could it be that the store is a corrupt one? And that it stays corrupt > since it doesn't get chance to cleanup... > > Cheers > Niclas _______________________________________________ qi4j-dev mailing list [email protected] http://lists.ops4j.org/mailman/listinfo/qi4j-dev
