Open rotten opened 8 years ago
@rotten could you take a thread-dump (kill -3
A thread dump of the shell or the database?
Possibly both as I don't know where the hanging actually occurs.
But if you run shell with "-path /path/to/graph.db" then only the shell.
I've been swamped since I reported this issue and haven't had a chance to experiment with the bigger database again. Meanwhile I have noticed that sometimes it is neo4j-shell that gets stuck. I've had this happen several times. When you try to connect with neo4j-shell, it just hangs indefinitely, even though otherwise the database appears to be functioning normally (and is mostly idle). If you bounce the database, neo4j-shell connects immediately. I'm wondering if my graphml export attempts were related. My bigger database is in such a state at the moment. I'll try getting stack traces of the plain neo4j-shell getting completely stuck and share those after its current round of data processing finishes. (It looks like it will take a couple more days to finish.)
@rotten did you try it with a shell that uses the on-disk files with -path data/graph.db
??
I hadn't tried it with that option. I was trying to connect via host/port. I just tried it, and it looks like I have to shut down the database server to use that option.
ERROR (-v for expanded information):
Error starting org.neo4j.kernel.impl.factory.CommunityFacadeFactory, /data/neo4j_db/graph.db
java.lang.RuntimeException: Error starting org.neo4j.kernel.impl.factory.CommunityFacadeFactory, /data/neo4j_db/graph.db
at org.neo4j.kernel.impl.factory.GraphDatabaseFacadeFactory.newFacade(GraphDatabaseFacadeFactory.java:143)
at org.neo4j.kernel.impl.factory.CommunityFacadeFactory.newFacade(CommunityFacadeFactory.java:43)
at org.neo4j.kernel.impl.factory.GraphDatabaseFacadeFactory.newFacade(GraphDatabaseFacadeFactory.java:108)
at org.neo4j.graphdb.factory.GraphDatabaseFactory.newDatabase(GraphDatabaseFactory.java:129)
at org.neo4j.graphdb.factory.GraphDatabaseFactory$1.newDatabase(GraphDatabaseFactory.java:117)
at org.neo4j.graphdb.factory.GraphDatabaseBuilder.newGraphDatabase(GraphDatabaseBuilder.java:185)
at org.neo4j.shell.kernel.GraphDatabaseShellServer.instantiateGraphDb(GraphDatabaseShellServer.java:210)
at org.neo4j.shell.kernel.GraphDatabaseShellServer.
I successfully used export-graphml without cypher a couple of weeks ago to dump a 40G (or so) neo4j database to graphml. (on a 32G system) It takes a while. It missed a few
<key />
definitions, but they were easy to add back to the XML file afterwards.I successfully used export-graphml with a
match
to export a subgraph on a smaller database (a few hundred thousand nodes and maybe 2M relationships). It starts writing the file within a couple of minutes.However, when I try that same export-graphml command to export a subgraph from the bigger database, it hangs (presumably) forever. I let it run nearly 24 hours with nothing - no errors, no changes to console.log, messages.log, nor the screen. It doesn't even look like it is consuming any cpu (although the process is still there). It creates a file of size Zero as soon as I start it, but never writes anything.
FWIW, import-cypher -o immediately starts writing a csv and seems to work. I suspect export-cypher would as well, but I haven't tried that because it mangles the integer properties. (issue #80)