Search Mailing List Archives
[bioontology-support] help loading rxnorm to ncbo virtual appliance
Paul R Alexander
palexander at stanford.edu
Tue Aug 7 18:52:47 PDT 2012
I'm looping Lian in to this conversation to see if he has any insight
(Pradip has moved on to another project, so Lian is the contact at Mayo
now). One of the things I can think to check is disk space. Is it
possible that it's filling up the space you have available on the
Lian -- any other ideas?
On Tue Aug 7 04:52:04 2012, Gardner, Gregory Allen wrote:
> Please help -- after many different attempts to load RXNORM into the NCBO virtual appliance, we are at a dead end (see below for errors and setup). Do you know what might be going wrong, or do you have a virtual appliance with RxNORM pre-loaded?
> Our initial error on attempting to load the RxNorm ontology (downloaded from http://rest.bioontology.org/bioportal/ontologies/download/46395) was a Java NullPointerException on the parse ontologies screen of the bioportal admin.
> We're running the appliance on version 11.10 of Ubuntu in VirtualBox. Due to our local network, we set the appliance to run with a host-only network adapter, rather than the default bridged adapter.
> Since the vm ip is then assigned by VirtualBox and cannot connect to an external network, we modified the rails bioportal configuration, setting the local_ip variable to the static ip assigned by virtualbox.
> Initially, we tried running with 4 gb of ram, increasing to 6.5 gb (the maximum our system can handle) after the initial failure. When running the vm at 6.5 gb of ram, we configure tomcat to use up to 5gb of ram by editting /etc/tomcat6/tomcat6.conf, changing the java opts from -Xmx2G to -Xmx5G.
> We load the local copy of rxnorm according to the instructions given to us by Pradip. Initially, a zip file with a dummy RRF file is loaded through the new ontology screen of bioportal (../ontologies/new). The folder where the appliance is placing the ontology files is then determined, and the local rxnorm RRF files are copied to this directory.
> The status of the ontology in the appliance at this point is "1", waiting to parse. We then tell bioportal to parse the new ontology at the bioportal admin parse ontologies screen. After clicking run, the ontology status changes to "2", parsing. Bioportal continues to parse the ontology for ~1-2 hours, at which point the above error is seen. Running at
> 4 gb of ram, the virtual machine also printed an "out of memory" error. At 6.5 gb, the process did not complete and eventually caused the host system to lock. Perhaps we do not have the memory required to load rxnorm ourselves?
> During various attempts to load the ontology, I occasionally also saw an error to the effect of "LexBIG ontology failed to load, exitcode=NON-RECOVERABLE" soon after (~10-20 seconds) telling the appliance to parse the new ontology. This seemed to occur when a previous attempt at loading RxNorm had failed and I tried to load it again under a different name without reloading a fresh copy of the appliance into VirtualBox.
> Please let me know if I can provide any further information.
> Thank you for your help,
> Greg Gardner
> Software Developer
> Department of Biomedical Informatics
> University of Pittsburgh
> bioontology-support mailing list
> bioontology-support at lists.stanford.edu
More information about the bioontology-support