I want to process many files on many database instances. Instances are on various computers, one computer may host up to 4 instances.
A process is run on an instance via a bash shell :
process-on.sh <file_to_process> <instance_name>
process-on.sh knows how to access the instance (server, script-name, and so on) from the instance_name. The same unix client is used to run processes for all instances on one host, just the remote script name varies.
I would like instances act as a pool of processors, and want to use gnu-parallel as a distributor of files between instances.
I would like to use gnu-parallel, but it is mainly to distribute between computers, not isntances... Maybe there is another tool which can work with pools...
Any hints will be much appreciated.
Best, Christophe