with our SeisJet Data Server software
How does it look like?
The SeisJet Seismic Data Server solution includes
the following components:
- Data Server daemon.
- Input/Output modules (SJ Input
and SJ Output) integrated into the processing
The Data Server is a standalone
program (daemon) that works solely with the seismic datasets of the
processing system. It is capable of quick seismic data reading and
sorting, and transfers the data to a number of copies of a
processing flow in a high performance manner.
The data from the Data Server
is introduced to the processing system batch through the input
module (SJ Input) that comes along with the
Data Server. The module initializes a new job in the Data
Server and feeds the processing flow with the data. To provide a
parallelization of the processing flow, the SJ
Input module can start a specified number of copies of the
same processing batch (“secondary” flows) on the different nodes of
the computational cluster (or on the same node for a SMP machine).
Each copy of the SJ Input module in the secondary
flows will also subscribe to the Data Server and
will feed each copy of the flow with the data. The output of the
flows are transferred back to the Data Server by
the SJ Output module. The Data
Server cares of parallel processing of the requests and
consistent data distribution between the flows, as well as
collects the result of the processing to an output
*ProMAX - is a
registered trademark of Landmark Graphycs Corporation.
declared speed-up estimation is based on our own tests as well as
some tests of our clients, executing conventional processing flows
on cluster computers. Actual speed-up may vary, depending on the
hardware configuration and software settings.