Database Reference
In-Depth Information
SRM Interface Layer
SRM Request Management Layer
SRM
File System
Interface Layer
SRM
Storage
Interface Layer
SRM
SRM
file System
Adaptation Layer
SRM
Storage
Adaptation Layer
File System
Storage System
Figure 3.3
An architectural diagram of a modular design of an SRM.
In general, SRMs that front MSSs can be adapted to work with various file
systems. The SRM developed by LBNL (called BeStMan) was adapted to front
HPSS, the NCAR MSS, L-Store, and Lustre. This is shown in Figure 3.3. It
was also adapted to provide an SRM gateway to the XROOTD
20
system used
by HEP projects. Similarly, StoRM developed in Italy fronts GPFS. There is
also a version of SRM developed in Taiwan to front SRB.
3.3 SRM Functionality
3.3.1 Background
The SRM technology was inspired by the emergence of the grid and the ability
to use distributed compute and storage resources for common tasks. The grand
vision of the grid is to provide middleware facilities that give a client of the
grid the illusion that all the compute and storage resources needed for their
jobs are running on their local system. This implies that a client only logs in
and gets authenticated once and that some middleware software figures out
where the most ecient way to run the job is; reserves compute, network and
storage resources; and executes the request. Initially the grid was envisioned
as a way to share large computing facilities by sending jobs to be executed
at remote computational sites. For this reason, the grid was referred to as a
computational grid
. However, very large jobs are often data intensive, and in
such cases it may be necessary to move the job to where the data is located
