Jaroslaw Cwiklik wrote:
Jörn, can you clarify the use case? Do you think that the uima AS deployment
descriptor should be extended to point to a pear descriptor?
Yes that would be nice.
My use case is that we have a small analysis cluster which
runs on a few machines. The cluster should be capable to run
multiple UIMA AS applications.
Has IBM experience with running UIMA AS on a cluster ?
Right now I still try to figure out how to setup UIMA AS in away
it can run on a cluster. Hadoop style start/stop scripts for all nodes
would be really
nice and a rsync script to copy over modifications made to one
node to the other nodes. Though maybe a JMX management application wold
be even better.
My AAE resources are hosted on a http server and updated frequently. To get
them on the servers I use a script which downloads the resources with wget.
Is there a way to make UIMA AS to re-initialize the running AAE instances to
load the updated resources ?
Jörn
