PROJECT TASKS
The PROGRESS project consisted of 10 grant funded tasks and an additional one involving development
of testbed applications.
-
Task 1.1. Service Management System for the Grid Environment
The task goal was to develop and deploy a system for managing resources and services of a computing
grid environment. The main paradigm of this system was to allow efficient execution of users' computing
jobs and efficient management if resources in a distributed and complex grid environment. The
PROGRESS resource management system took a form of a Grid Resource Broker, which allows the users to
specify any necessary attributes and resource requirements for their applications and to define work-flow
between tasks. Thanks to that it is possible to construct advanced computing experiments,
to specify preferences for the job execution process, to monitor the results and the execution of tasks, etc.
The Grid Resource Broker uses the Extended Resource Specification Language
(XRSL) for grid job definitions.
The experiences gained during the PROGRESS Grid Resource Broker design were a great base for construction
of the next generation Grid Resource Management System (GRMS) in the scope of the
GridLab project.
The GRMS is set to replace the PROGRESS GRB in the PROGRESS HPC Portal environment by the end of this year.
More...
-
Task 1.2. Increasing Security in a Grid-Portal Infrastructure through Active Intrusion
Detection
The main task goal was to design and develop an intrusion detection system for grid-portal environments.
The developed system, VALKYRIE, which is based on experiences gained from a reasearch project VALIS
realized previouslu in PSNC, is capable of real-time intrusion detection. VALKYRIE is a combination
of mechanisms known from IDS systems as well as basic mechanisms of reaction to detected events.
The system secures control and monitoring of selected system for immediate detection of any
attack trials. More...
-
Task 1.3. Checkpointing for Solaris Operating System
The first goal of this task was to optimize and adapt one of available library packages for user level
checkpointing. The library chosen for adaptation was libkckp, due to its availability
in the source code distribution. The aim of library optimization was to adapt its functionality
to specific commands of the processor used in Sun servers and to optimize the memory access method
to shorten the time necessary to make a checkpoint. The second goal of the task was to extend
the operating system with the kernel level checkpoint function. More...
-
Task 2.1. Data Management and Processing, and Database Interfaces
The task goal was to design a distributed system allowing to efficiently manage
large amounts of scientific data. This was achieved by designing the Data Management System (DMS).
DMS main function is to store and provide access to data. DMS also provides possibilities to describe
data with metainformation, which facilitates data search. The data can be stored on various media
like for example file systems, archivers and relational databases. The DMS Proxy module enables
access to external scientific database. The whole system was based on Web Services technology.
More...
-
Task 2.2. Portal Services and Application Management
The task goal was to build a portal access environment to grid resources, which would enable
management of grid-enabled applications and submission of computing experiments for execution
on the grid. The task introduced a new layer in the grid-portal envitonment, the Grid Service Provider,
which groupped the data analysis functionality performed by previously known portal services within
separate modules expossed as Web Services. The portal task in this architecture is to collect the input
from the user, translate the HTTP request into invocation of proper WS methods, incorporating the collected input
into these method invocation and presenting the results obtained from the services to the user.
More...
-
Task 2.3. Application Results Visualization
The main goal of this task, coordinated by
ACC "Cyfronet" AGH, was to analyze the capabilities of
various visualization techniques in the context of grid environments. Based on the gathered knowledge
a practical implementation of three visualization modules for bioinformatic applications (see Task 4)
was conducted. More...
-
Task 2.4. Work and Access Environment for Mobile Users in a Grid Infrastructure
The goal of this task was to create tools allowing to restore exactly the same user workplace
environment despite of the user workstation current location. Traditionally, the user utilizing
grid resources from multiple different workstations had seperately defined workplace environment
on each of them. This included access paths to applications, grid access certificates or personalized
graphical environment look-and-feel. That is why the access to grid resurces could be different on each
of workstations due to local workstation configuration. The Grid Desktop, which first version was
prepared within the PROGRESS project, is a user environment, which is available from any place via a web
browser. The user has a possibility to configure the workplace environment (desktop) to suit his/her activity
needs. The PROGRESS Grid Desktop was a base for further research and extensions within the
Crossgrid project.
More...
-
Task 3.1. Contruction of the computing cluster infrastructure
The goal of this task was to construct and install the testbed computing cluster, consisting of three
24-CPU Sun Fire 6800
servers. Two of these servers were installed in Poznań and one in Kraków. Additionally,
two Sun Fire V880 servers, one at each
location, were installed as the infrastructure for the Data Management System, as well as a front-end server,
a Sun Fire 280R (in Poznań),
which became the platform for the computing portal.
-
Task 3.2. Integration of the cluster in the PIONIER optical internet network
The goal of this task was to prepare dedicated WDM channels of the
PIONIER optical network to secure
high throughput for data transfers between computing cluster elements. The integration of the cluster in the
optical network environment allows to guarantee fast, comfortable and secure computing tasks maintenance
in the realized grid-portal environment.
-
Task 3.3. Integration of software for the computing cluster
The goal of this task was to analyze the capabilities of the software enabled for the realization
of project tasks. The software included commercially available packages like
Oracle 9i relation database,
Sun HPC Cluster Tools,
Sun Grid Engine,
iPlanet (later Sun ONE, currently
Sun Java System) software line, as well as software obtained through open source or GNU licences,
for example Globus Toolkit. From this set of tools
those that were chosen as environments for specific modules developed within other tasks got installed
and integrated into the base software infrastructure for the PROGRESS grid-portal environment.
-
Task 4. Testbed bioinformatic applications
The designed and deployed grid-portal environment architecture and the tools prepared to
build it were tested by three bioinformatic applications. These applications have been
developed by the Laboratory of Bioinformatics of the Institute
of Bioorganic Chemistry Polish Academy of Sciences and have been: