Re: Looking for scientific DB to manage physics sim. output
Posted in 1996
Topics: Storage & Space Management, Clustering, Grid & MACH11, Cloud, Docker & Containers
Looks like something the Illustra database might be able to handle, and later the Informix Universal server. Look at www.illustra.com or www.informix.com Mark Miller <miller@viper.llnl.gov> wrote: :I am interested in finding a commercially available database :for managing sceintific data which can meet the following require- :ments... :1. Can support large class of scientific data including struc- :tured and unstructured FEM grids, variables defined over these :grids centered at nodes, zones, faces or edges and variables defined :over just a subset of nodes, zones, faces, edges. :2. Accepts a number of different storage orders for different :types of meshes and mesh variables through the API. In other :words, if I store an unstructred mesh composed of tetrahedra and :hexhedra, I would like the SDB to accept any ordering of the layout :of the nodes making up the tetrahedra and hexahedra. Alternatively, :if I have a structured mesh, I would like to be able to store it x- :major, y-major, z-major, interleaved and or chunked. :3. Supports decompositions of a data among a number of :processors working in parallel, such as on an MPP. :4. Provides numeric-architecture independent access to the :data. In other words, you can write it on a cray and read it on an Sgi. :5. Performs within 80% of raw, C or Fortran binary I/O, :when writing/reading on the same numeric-architecture machine. :6. Can operate in stand-alone and client/server modes. :7. Stores the data in "containers" (probably files but I am not :sure) that can be moved across filesystems and the network easily. :8. Supports spatial data queries such as, "read all zones :around the point (x,y,z) or read all zones within the box (x0,y0,z0)- :(x1,y1,z1)" :9. Does not cost an arm and a leg :10. Stores structured meshes with the efficiency they pro- :vide. :Does anyone know of a product supporting a majority of :these requirements? :reply to miller@viper.llnl.gov :-- :Mark C. Miller, Lawrence Livermore National Laboratory :email: miller@viper.llnl.gov :phone: (510)-423-5901 :fax: (510)-424-2709 Nils.Myklebust@ccmail.telemax.no NM Data AS, P.O.Box 9090 Gronland, N-0133 Oslo, Norway My opinions are those of my company