Fail to preallocate additional space for store file

Strange error in debug log.

Caused by: org.neo4j.exceptions.UnderlyingStorageException: java.io.IOException: Fail to preallocate additional space for store file at: neostore.relationshipstore.db. Requested file size: 2550136832. Call error: ErrorCode=27, errorMessage='File too large'
at org.neo4j.kernel.impl.store.CommonAbstractStore.updateRecord(CommonAbstractStore.java:941) ~[neo4j-record-storage-engine-4.4.4.jar:4.4.4]*
at org.neo4j.internal.recordstorage.NeoStoreTransactionApplier.updateStore(NeoStoreTransactionApplier.java:214) ~[neo4j-record-storage-engine-4.4.4.jar:4.4.4]*
at org.neo4j.internal.recordstorage.NeoStoreTransactionApplier.visitRelationshipCommand(NeoStoreTransactionApplier.java:100) ~[neo4j-record-storage-engine-4.4.4.jar:4.4.4]*
at org.neo4j.internal.recordstorage.Command$RelationshipCommand.handle(Command.java:260) ~[neo4j-record-storage-engine-4.4.4.jar:4.4.4]*
at org.neo4j.internal.recordstorage.TransactionApplierFacade.visit(TransactionApplierFacade.java:57) ~[neo4j-record-storage-engine-4.4.4.jar:4.4.4]*
at org.neo4j.internal.recordstorage.TransactionApplierFacade.visit(TransactionApplierFacade.java:32) ~[neo4j-record-storage-engine-4.4.4.jar:4.4.4]*
at org.neo4j.kernel.impl.transaction.log.PhysicalTransactionRepresentation.accept(PhysicalTransactionRepresentation.java:82) ~[neo4j-kernel-4.4.4.jar:4.4.4]*
at org.neo4j.kernel.impl.api.TransactionToApply.accept(TransactionToApply.java:153) ~[neo4j-kernel-4.4.4.jar:4.4.4]*
at org.neo4j.internal.recordstorage.RecordStorageEngine.apply(RecordStorageEngine.java:505) ~[neo4j-record-storage-engine-4.4.4.jar:4.4.4]*

Any Idea?

@kcw
I presume this is Neo4j 4.4.4? correct? if not can you provide further specifics.

Does your file system have ample free disk space, i.e. from a linux command run df -h

Yes. 4.4.4

I have plenty of disk space.

image

@dana_canzano Any Ideas?

@kcw

hard to say,
is there any context about this error?
Is this error encountered at time of startup? at time of online-ing the database? at time of backup ?

does linux command

ulimit -a

report an unlimited file size?

and if i forceibly set a exceeding low limit as a test I can reproduce a File size limit exceeded(core dumped), though not exactly the same but similar.
And if so this doesnt appear to be a Neo4j failure per say, but more so behaving as expected given the low file size limit of linux

@kcw

any progress? was the OS limiting the max file size?

Yes, it was!!!

Thanks.

1 Like

@kcw
great to hear it wasnt directly a Neo4j limitation/issue