m3l0n-sq commented on issue #3488: URL: https://github.com/apache/jena/issues/3488#issuecomment-5114613802
I ran into this issue on Jena 6.1.0 as well. A daily compaction cron job at 1am UTC turned out to be stuck since 23rd to 28th July. Same symptoms as author, `org.apache.jena.dboe.transaction.txn.TransactionException: Not in a transaction`. Note: RAM and disk were plentiful at this moment. `/$/tasks` endpoint shows a list of compaction jobs queued up. On 29th July, the dataset ballooned from 7gb to 196gb with some heavy usage, disk were at 100% usage. This was when we face the error `java.lang.InternalError: a fault occurred in an unsafe memory access operation`. This was Claude's explanation: > java.lang.InternalError: a fault occurred in an unsafe memory access operation is the JVM's translation of a SIGBUS while touching a memory-mapped file. TDB2 writes its B+Tree indexes through MappedByteBuffers, and when the filesystem can't allocate a backing block for a mapped page — i.e. the disk is full — the kernel delivers SIGBUS and the JVM surfaces it as exactly this error. The stack confirms it: the fault is in BufferBase.copy / BPTreeNode.split while writing the new, compacted copy of the database (CopyDSG.copy → StorageTDB.add). -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
