Managing our Grid Node, Involvement under Trust Fabric, External Collaboration, & In-House Projects Adeel-ur-Rehman on behalf of Advanced Scientific Computing Group (ASC)
Scheme of Talk Grid Computing NCP-LCG2 (T2_PK_NCP) Certification Authority PK-GRID-CA In-House HPC Framework NCP Cluster Projects with CERN CMS Collaboration Software Development & Support Assistance for EHEP fellows 6/10/2016 ASM
Grid Computing 6/10/2016 ASM
What are Grids? A Grid is: Special form of distributed computing o Computing and storage resources are distributed over several locations called sites o Sites are typically connected via WAN links o Site normally has a local-area network which itself has distributed computing and data storage resources Coordinated resource sharing and problem solving in dynamic, multi-institutional Virtual Organizations (VO) o A VO is a collection of users sharing similar needs and requirements in their access to processing, data and distributed resources and pursuing similar goals. 6/10/2016 ASM
Grid in terms of VOs 6/10/2016 ASM
LHC & WLCG The Large Hadron Collider (LHC) – the huge particle accelerator: is constructed at the European Laboratory for Particle Physics (CERN), at Franco-Swiss border near Geneva, Switzerland. is the world’s largest and most powerful particle accelerator. the experiments using it are generating very large amounts of data (in Peta-Bytes / year). 6/10/2016 ASM
LHC & WLCG The job of the Worldwide LHC Computing Grid Project (WCLG) is to prepare the computing infrastructure for the: Simulation processing and analysis of LHC data for all initial four of the LHC collaborations: o ALICE, ATLAS, CMS, and LHCB The processing of this data requires enormous computational and storage resources. 6/10/2016 ASM
WLCG at NCP Pakistan initiated collaboration with CERN for CMS experiment in 1990s. Consequently, the effort to bring Pakistan on the WLCG map as a Grid Node also started. A Grid Technology Workshop was organized by NCP from October 20-22, The first ever testbed was deployed during the workshop for tutorial. 6/10/2016 ASM
WLCG at NCP Now NCP is an official Grid Node in Pakistan. Tested & Certified by the Grid Deployment Team (dteam) at CERN first in South Asia and fifth in Asia Since then, we have been appearing on the Grid Operations Centre (GOC), as well as GStat websites: &object_id=101&grid_id=0 &object_id=101&grid_id=0 LCG2/#fragment-1 LCG2/#fragment-1 The node has been constantly upgraded w.r.t. WLCG middleware and versions switching b/w: Globus gLite EMI 1, 2 (now heading towards EMI3) 6/10/2016 ASM
NCP on WLCG Globe 6/10/2016 ASM T2_PK_NCP
6/10/2016 ASM Installed Resources Physical CPU106 Logical CPU /Cores524 KSI2K1591 Disk (TB)260 Network (Mbps)155 T2_PK_NCP Site
Trust Fabric Involvement 6/10/2016 ASM
Grid Security Infrastructure (GSI) The GSI in WLCG enables secure authentication and communication over an open network. GSI is based on public key encryption, X.509 certificates, and the Secure Sockets Layer (SSL) communication protocol. Extensions to these standards have been added for single sign-on and delegation. So, in order to access the WLCG framework, one must have a digital certificate signed by a trusted Certification Authority (CA). 6/10/2016 ASM
Certification Authority (CA) A CA is an executive body which issues certificates for users, programs, and machines. A digital certificate is an electronic "credit card" that establishes our credentials when doing business or other transactions on the Web. It uses Public Key Infrastructure (PKI) Enables users of a public network to exchange data securely using a public/private cryptographic key pair obtained and shared through a trusted authority (CA). 6/10/2016 ASM
PK-GRID-CA NCP is itself a CA (non-commercial) which provides X.509 certificate (user/host) to support the secure environment in grid related projects NCP produced the first Certificate Policy and Certification Practice Statement (CP-CPS) document in December Reviewed by several members of European Grid Policy Management Authority (EU-Grid-PMA) that works under the umbrella of International Grid Trust Federation (IGTF). IGTF also takes care of APGridPMA (for Asia Pacific), and TAGPMA (for the whole America) 6/10/2016 ASM
PK-GRID-CA Three revisions were made which resulted from comments and suggestions by PMA members. The CA was presented in September 2004 in the 2 nd meeting of the EU-Grid-PMA held in Brussels. NCP was formally approved by the EU-Grid-PMA as a Certification Authority. PK-Grid-CA had started operations since then. First Certification Authority in Pakistan. For more information: 6/10/2016 ASM
6/10/2016 ASM PK-GRID-CA The routine task comprises of (but not limited to): issuing user/host certificates for our subscribers generating Certificate Revocation Lists (CRLs) revoking certificates when needed signing root key for our CA when due managing CA web portal for handling user requests maintaining users’ records, necessary correspondence and required cryptographic data pertaining to our root certificate recording each and every interaction of the CA offline server monitoring CA premises maintaining and updating our Certificate Policy/Certification Practice Statement (CP/CPS) document as required etc.
6/10/2016 ASM PK-GRID-CA Statistics Current PK-Grid-CA stats: Total Certs Issued : 345 o User Certs : 204 o Host Certs : 141 o Certificates Expired : 206 o Certificates Revoked : 54 o Active Certificates : 85 For more info;
6/10/2016 ASM High Performance Computing Cluster
NCP Cluster To provide our scientists the access to High Performance Computing resources for running simulation codes to model their research problems. Over 30 researchers from all over the country have been facilitated with our cluster resources. 6/10/2016 ASM
NCP Cluster 6/10/2016 ASM
NCP Cluster This cluster has been involved to conduct research and development under diversified areas of study like Ion Channeling, Multi-Particle Interaction, Space Physics, Weather Forecasting, Density Functional Theory (DFT) etc. 6/10/2016 ASM
NCP Cluster Hardware Resources and Environment: SunFire Intel Xeon Machine 16 GB RAM 8 cores/node with 4 computational nodes Scientific Linux CERN 5.3 OS f77, g77 (gfortran), gcc openPBS (for batch processing) mpich p1 (for parallel processing) 6/10/2016 ASM
Software Development/Testing with CMS (CERN) Collaboration 6/10/2016 ASM
RPC Construction DB Designing, development and testing of ConstructionDB wih Resistive Plate Chambers (RPC) group under CMS. 6/10/2016 ASM RPC Construction DB Barrel DB Endcap New Upgrade
RPC Construction DB The software is being built in terms of various levels of Quality Controls (QCs) leading to different development phases. QC1 (Chamber components validation) QC2 (GAPS certification) QC3 (Chambers test) QC3_A Chambers assembly validation (Implemented on production server) QC3_B Tests on Cosmic stand (Implemented on production server) QC4 Chamber reception & acceptance from sites (Implemented on production server) Super-module assembly and final test before shipment to P5 (Testing Phase) QC5 Installation and commissioning a P5 (Still to be done) 6/10/2016 ASM
Testing of DQM Sequences Testing performance & goodness of offline Data Quality & Monitoring (DQM) modules within CMSSW project. Looking for compile-time and run-time failures Observing memory consumption fluctuation Notifying the corresponding persons at CERN about our findings in order to decide for the integration of a particular piece of DQM code within CMSSW 6/10/2016 ASM
Testing of DQM Sequences The exercise has been carried out over various release cycles of CMSSW like 5_2_X, 5_3_X, 6_0_X, 6_1_X, 6_2_X and currently 7_0_X. Other than that, we are also involved in adopting this workflow as per need of the outer environment such as: Integration of the test suite within CMSSW CVS -> Git (Revision Control Systems) Integration with automated build systems (Jenkins) 6/10/2016 ASM
Miscellaneous 6/10/2016 ASM
In-House Development and Support Employee Salary System Finance Ledger System (FLS) Online Leave Application System (OLAS) Library Information Management System (LIMS) Hardware Inventory Management System (HIMS) or Network Resource Management System (NRMS) Store Management System Transport Requisition System Redmine Ticketing System ISS Conference Registration System 6/10/2016 ASM
Training & Assistance Following types of facilities are offered to our EHEP students for their research work: Periodic conduction of courses/tutorials: o C++ Programming Basic Advanced o Python Programming Providing help in describing the source code of analysis software such as CMSSW, ROOT etc. Involvement in troubleshooting the grid site operations on demand. 6/10/2016 ASM
References “WLCG Node in Pakistan – Challenges & Experiences”, by Sajjad Asghar, Usman Ahmad Malik & Adeel-ur-Rehman, Managed Grids and Cloud Systems in the Asia-Pacific Research Community 2010, pp , Springer.Managed Grids and Cloud Systems in the Asia-Pacific Research Community “Establishment of Public Key Infrastructure in Pakistan”, by Sajjad Asghar, Usman Ahmad Malik & Adeel-ur-Rehman 8th National Research Conference, SZABIST, Islamabad 6/10/2016 ASM
6/10/2016 ASM