I have a big graph search algorithm which has implemented through Neo4j
java API (Using as a Neo4j Procedure). I am trying to use threads
(thread-pool) in order to improve the performance of the algorithm. The
algorithm is totally a read-only but each thread needs to be a transaction
in Neo4j to satisfy the ACID properties.
try (Transaction tx= db.beginTx()) {
......
tx.success();
}
Since this is a transaction, I think a node can be accessed by a single
thread at once. Please correct me if I am wrong.
- Is there a way to avoid this and access the same node by multiple
threads at once because the algorithm performs read-only operations?
(Otherwise, I cannot achieve significant performance improvement)
At the end of the execution, I have to collect all the results from threads
and use *Stream<Object>* to retrieve search results. So, threads are not
totally executed independently because a common data structure is needed to
collect all thread results after threads execution. To address that, I have
used *Callable<Object>* interface. The problem with this approach is, it
waits to all thread results sequentially. Some threads which start later
can be finished the execution earlier compared to some other older threads.
Since results are collected sequentially, already finished threads have to
wait until get their chance.
- Is there a way to overcome this issue or any alternate for *Callable
*interface
that we can collect results to a common data structure as soon as threads
complete their work?
--
You received this message because you are subscribed to the Google Groups
"Neo4j" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/neo4j/ff75b166-bcce-4b89-ac1c-2a4102ad90b5%40googlegroups.com.
