SlapOS is still a proof of concept, a working proof of concept, yet a proof of concept. The SlapOS project is lead by ViFiB, the inventor of SLAPgrid based on scientific ideas imagined by Prof.Christophe Cérin of Paris University (XIII). Nexedi has contributed ERP5 Cloud Engine to the SlapOS project.
ViFiB is conducting currently an alpha phase to experiment SlapOS with kvm instance and TioLive instances as first candidates. A beta phase will soon follow. It is a good occasion for any company willing to operate a SaaS, a PaaS or a IaaS to contact ViFiB and contribute real use cases and real business cases.
As part of the COMPATIBLE collaborative R&D project, Nexedi will contribute to SlapOS various extensions to make it standard and compatible with more applications. Again, it is now a good time to formulate what you would like to see happening in SlapOS: documentation, specification of SLAP protocol, interfaces to remotely control SlapOS, AppStore interfacing. By formulating it now, ViFiB and Nexedi can take R&D directions which can help you in the future. Please keep in mind that both ViFiB and Nexedi are willing to help you cook your own recipes for your own software solution.
In addition, Nexedi will use SlapOS to conduct tests and benchmarks on NoSQL storages such as KumoFS, NEO, etc. SlapOS greatly simplifies the packaging and the elastic deployment of software across a cloud. It is thus a great tool to run tests in parallel over hundreds or thousands of servers.
On the R&D front line, implementation of SLA in SLAP is still an open question. According to SlapOS philosophy it should be handled by the recipe or the appliation. Yet, some minimal and simplistic SLA specification (ex. “on the same server”, “on the same LAN”) and exception handling (ex. “no resource available”) which seems to be necessary remains to be designed and implemented. Help is welcome.