"Page type mismatch" after removing index about zoodb HOT 7 CLOSED

Scheduled for 0.5.2 / 0.6.0

from zoodb.

Hi there, we are running exactly into this problem. So is there any solution, because here I can't see anyone? Or are we supposed to wait until version 0.6? If we have so, what is the expected release time?

from zoodb.

Ahh, it was fixed from version 0.5.1 to 0.5.2, isn't it? We still use 0.5.1.

from zoodb.

Yes, this particular issue was fixed in 0.5.2. If you are removing indexes, this may very well be fixed by using 0.5.2. Let me know if it does not.

Btw, I am currently preparing a 0.6.0, it will mainly fix issues with multiple concurrent PersistentManagers: #114 and #103 .

from zoodb.

Sorry, but unfortunately we have to reopen the issue again. After recompiling our tool with ZooDB 0.5.2 we still get the "Page type mismatch" error. How can we proceed, should I provide you the corrupted database file?

The exception stack trace is:
javax.jdo.JDOFatalInternalException: Page type mismatch, expected 4/OID_INDEX (tx=0) but got 0 (tx=0). PageId=718102
at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
at sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:62)
at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:45)
at java.lang.reflect.Constructor.newInstance(Constructor.java:423)
at org.zoodb.internal.util.ReflTools.newInstance(ReflTools.java:211)
at org.zoodb.internal.util.DBLogger.newEx(DBLogger.java:108)
at org.zoodb.internal.util.DBLogger.newEx(DBLogger.java:99)
at org.zoodb.internal.util.DBLogger.newFatalInternal(DBLogger.java:162)
at org.zoodb.internal.util.DBLogger.newFatalInternal(DBLogger.java:158)
at org.zoodb.internal.server.StorageReader.readHeader(StorageReader.java:334)
at org.zoodb.internal.server.StorageReader.seekPage(StorageReader.java:104)
at org.zoodb.internal.server.StorageReader.seekPageForRead(StorageReader.java:80)
at org.zoodb.internal.server.index.AbstractPagedIndex.readPage(AbstractPagedIndex.java:173)
at org.zoodb.internal.server.index.AbstractPagedIndex.readRoot(AbstractPagedIndex.java:160)
at org.zoodb.internal.server.index.PagedUniqueLongLong.(PagedUniqueLongLong.java:57)
at org.zoodb.internal.server.index.PagedUniqueLongLong.(PagedUniqueLongLong.java:52)
at org.zoodb.internal.server.index.IndexFactory.loadUniqueIndex(IndexFactory.java:105)
at org.zoodb.internal.server.index.PagedOidIndex.(PagedOidIndex.java:190)
at org.zoodb.internal.server.SessionManager.(SessionManager.java:148)
at org.zoodb.internal.server.SessionFactory.getSession(SessionFactory.java:88)
at org.zoodb.internal.model1p.Node1P.connect(Node1P.java:74)
at org.zoodb.internal.Session.(Session.java:112)
at org.zoodb.jdo.impl.PersistenceManagerImpl.(PersistenceManagerImpl.java:101)
at org.zoodb.jdo.impl.PersistenceManagerFactoryImpl.getPersistenceManager(PersistenceManagerFactoryImpl.java:97)
at org.zoodb.jdo.ZooJdoHelper.openDB(ZooJdoHelper.java:54)
at org.zoodb.jdo.ZooJdoHelper.openOrCreateDB(ZooJdoHelper.java:81)

from zoodb.

Yes, the database file may be helpful.
However, it would be easiest if I had a program that reproduces the problem, or does it only happen with this database file?
Could you also describe how the database got into this state, e.g. were any special operations executed, concurrent sessions, ... .

Is this still compomics-utilities? Which branch?

from zoodb.

I created a new issue #122 to cover this problem.

from zoodb.