I am planning to use couchdb on a Linux VM with some limited local storage and a huge amount of NFS storage. What I would like to do is run one couchdb instance, but keep some databases on the local storage and some on the NFS.
-If the NFS goes down and the couchdb process enters the 'D' state, will I still be able to talk to couch over HTTP? -Furthermore, will I still be able to access databases on the local VM storage?
As I asked in the question, the couchdb process could go into the 'D' state for prolonged periods of time if it is performing an I/O operation over NFS. This state is NOT interruptible in any way.
Does couchdb account for this situation? Is there a strategy for using multiple couch processes?