Hi OpenAnzo Group,
If there is a NamedGraph with millions of triples stored on the server
and you would like to get a certain revision of this graph then you
do a
IAnzoGraph namedGraphRevision =
anzoClient.getNamedGraphRevision(graph.getNamedGraphUri(), new
Long(422));
The problem with this call is that you get an exception on the client
if the graph is too big. There is no exception on the server ( i.e. in
the /var/anzo.log file)
<--- snip Exception start --->
org.openanzo.common.exceptions.AnzoException: ErrorCode[16:16393]
[COMBUS_ERROR] No Response from Server for request
74dc5e10-294b-4b98-82eb-322d4d1cb7d1.
at
org.openanzo.servicecontainer.backends.JMSConnectionBackend.requestResponse(JMSConnectionBackend.java:
731)
at
org.openanzo.servicecontainer.services.proxy.jms.JMSModelServiceProxy.getNamedGraphRevision(JMSModelServiceProxy.java:
197)
at
org.openanzo.servicecontainer.services.proxy.jms.JMSModelServiceProxy.getNamedGraphRevision(JMSModelServiceProxy.java:
168)
at
org.openanzo.client.AnzoClient.getNamedGraphRevision(AnzoClient.java:
863)
at ....
<--- snip Exception end --->
In the documentation available at "
http://www.openanzo.org/projects/openanzo/wiki/AnzoClientDesign
" following is stated:
" getNamedGraphRevision - Return an in-memory copy of the given named
graph at the given revision. The returned graph may be modified in
memory, but these modifications have no effect on the named graph in
neither the repository nor the local replica"
Is there a way round the in-memory problem of big graphs?
Is it possible to replace a big graph with a earlier revision?
Thanks and all the best,
Andreas