Backup error on database with unused or corrupt data

[arikan]

The backup process throws the following error, most likely due to corrupt or unused data. This problem was posted as an issue earlier and hasn't been resolved.

command failed: Backup failed: Error starting org.neo4j.com.storecopy.ExternallyManagedPageCache$GraphDatabaseFactoryWithPageCacheFactory$1, /disk2/neo4j-backup/default_backup/temp-copy

github.com/neo4j/neo4j

Backup Error

opened 11:38PM - 21 Jul 18 UTC

arikan

We've been having a backup error, details below. This process used to work, nothing changed regarding neo4j settings. We'd appretiate your...

team-cluster

We did dump and load the database, and then cleaned up with store utils, however it resulted quite unexpectedly, destroyed all the relationships in the database and connected them to a single node with id 0. See the issue below:

github.com/jexp/store-utils

Store-utils release 3.5.20 destroys the graph by reconnecting all relationships to node id 0

opened 11:12PM - 14 Sep 20 UTC

arikan

We get a really weird behavior with store-utils (tried in release 3.5.20 and release 3.5.6). It removes all relationships --millions of...

So we are left helpless with a corrupt database, with no way to do proper backup (we have earlier backups that are ok but there are new data on production), no utils to clean. We use neo4j 3.5.5. We'd appretiate your suggestions on how to deal with these two issues? Thanks!

[dmtougaw]

We had the same issue with a remote backup throwing the following

Backup failed: Error starting org.neo4j.com.storecopy.ExternallyManagedPageCache$GraphDatabaseFactoryWithPageCacheFactory$1

We were able to solve it by changing user permissions. Seems like the script needs root privileges to handle Cache.