On Nov 16, 8:56 am, Asaf Hayman <[EMAIL PROTECTED]> wrote: > Is anyone familiar or aware of a successful enterprise class project > in Python to control and monitor a cluster of computers? > > As a part of a bigger project my company needs to build a cluster > management system. The aim of the system is to control and synchronize > applications. Namely, a central management will get events from the > remote applications and from a user and will react with commands to > the applications. > Existing tools don’t fit our requirements, so we will have to write it > ourselves. > We are currently pondering which programming language will best suite > us. The two major contenders are Python and Java. > > Some requirements: > 1. Cluster size is about 100 nodes. Only few applications per node. > 2. Database interface. > 3. Rate of messages: Few hundred messages per second at peak time. > 4. External web application for control (only few users) > 5. RPC between the management system to application written in c. > 6. The whole system will run within a LAN, and the required > responsiveness latency should be no more than 1sec. > > There are probably some more requirements which are common for cluster > management systems. > > We are currently looking for success stories of a similar projects > done in Python. It will be great if someone could point me to a > success story. It will be even better if I could talk with someone > which was involved in such a project. > > Thanks, > Asaf
Not sure if http://pypi.python.org/pypi/Fabric/0.0.3 is what you need or not; maybe that in concert with Nagios. -- http://mail.python.org/mailman/listinfo/python-list
