On 27/04/2021 09:59, Mikael Pesonen wrote:
That's our guess too, but would be nice to have some idea where to look
for the cause. Does Jena/Fuseki handle disk full situations without
corruption?
It should do (the transaction aborts) which is why I was asking.
Bulk loading, other than loader "basic" which is safe - depends exactly
when it happens in the process i.e. no guarantees.
Andy
On 27/04/2021 11.56, Andy Seaborne wrote:
In the original message,
There was shortage of disk space, hope the db is not corrupted.
What happened?
This is the only thing you've mentioned that relates to update.
Everything else is "read" and the fault occurred at an earlier time or
its an environmental factor (one mentioned file access permissions
e.g. another process is interfering with files).
Apr 12 12:30:55 solid java[22910]: [2021-04-12 12:30:55] Fuseki WARN
[346] RC = 500 : a fault occurred in a recent unsafe memory access
operation in compiled Java code Apr 12 12:30:55 solid
java[22910]: at
org.apache.jena.dboe.base.buffer.RecordBuffer.compare(RecordBuffer.java:192)
~[fuseki-server.jar:3.17.0]
so JDK ByteBuffer.get failed bu works almost always. It is likely to
be an environmental factor (the file system, background process
messing around, hardware issue).
Andy
