StoRM Architecture and Daemons

Slides:



Advertisements
Similar presentations
Jens G Jensen CCLRC/RAL hepsysman 2005Storage Middleware SRM 2.1 issues hepsysman Oxford 5 Dec 2005.
Advertisements

Data Management Expert Panel - WP2. WP2 Overview.
Data Management Expert Panel. RLS Globus-EDG Replica Location Service u Joint Design in the form of the Giggle architecture u Reference Implementation.
EGEE is a project funded by the European Union under contract IST Using SRM: DPM and dCache G.Donvito,V.Spinoso INFN Bari
Storage: Futures Flavia Donno CERN/IT WLCG Grid Deployment Board, CERN 8 October 2008.
– n° 1 StoRM latest performance test results Alberto Forti Otranto, Jun
GGF Toronto Spitfire A Relational DB Service for the Grid Peter Z. Kunszt European DataGrid Data Management CERN Database Group.
1 Exploring Data Reliability Tradeoffs in Replicated Storage Systems NetSysLab The University of British Columbia Abdullah Gharaibeh Matei Ripeanu.
Web-based Portal for Discovery, Retrieval and Visualization of Earth Science Datasets in Grid Environment Zhenping (Jane) Liu.
Tanenbaum & Van Steen, Distributed Systems: Principles and Paradigms, 2e, (c) 2007 Prentice-Hall, Inc. All rights reserved DISTRIBUTED SYSTEMS.
1 Exploring Data Reliability Tradeoffs in Replicated Storage Systems NetSysLab The University of British Columbia Abdullah Gharaibeh Advisor: Professor.
CERN, 29 August 2006 Status Report Riccardo Zappi INFN-CNAF, Bologna.
Unit – I CLIENT / SERVER ARCHITECTURE. Unit Structure  Evolution of Client/Server Architecture  Client/Server Model  Characteristics of Client/Server.
D C a c h e Michael Ernst Patrick Fuhrmann Tigran Mkrtchyan d C a c h e M. Ernst, P. Fuhrmann, T. Mkrtchyan Chep 2003 Chep2003 UCSD, California.
Introduction to dCache Zhenping (Jane) Liu ATLAS Computing Facility, Physics Department Brookhaven National Lab 09/12 – 09/13, 2005 USATLAS Tier-1 & Tier-2.
WNoDeS – Worker Nodes on Demand Service on EMI2 WNoDeS – Worker Nodes on Demand Service on EMI2 Local batch jobs can be run on both real and virtual execution.
Author - Title- Date - n° 1 Partner Logo EU DataGrid, Work Package 5 The Storage Element.
Hands-On Microsoft Windows Server Implementing Microsoft Internet Information Services Microsoft Internet Information Services (IIS) –Software included.
NA-MIC National Alliance for Medical Image Computing UCSD: Engineering Core 2 Portal and Grid Infrastructure.
Overview of Privilege Project at Fermilab (compilation of multiple talks and documents written by various authors) Tanya Levshina.
6/23/2005 R. GARDNER OSG Baseline Services 1 OSG Baseline Services In my talk I’d like to discuss two questions:  What capabilities are we aiming for.
Padova, 5 October StoRM Service view Riccardo Zappi INFN-CNAF Bologna.
David Adams ATLAS ATLAS distributed data management David Adams BNL February 22, 2005 Database working group ATLAS software workshop.
Data Transfer Service Challenge Infrastructure Ian Bird GDB 12 th January 2005.
Andrea Manzi CERN On behalf of the DPM team HEPiX Fall 2014 Workshop DPM performance tuning hints for HTTP/WebDAV and Xrootd 1 16/10/2014.
EGI-Engage Data Services and Solutions Part 1: Data in the Grid Vincenzo Spinoso EGI.eu/INFN Data Services.
AFS/OSD Project R.Belloni, L.Giammarino, A.Maslennikov, G.Palumbo, H.Reuter, R.Toebbicke.
1 Introduction to Active Directory Directory Services Uniquely identify users and resources on a network Provide a single point of network management.
BNL dCache Status and Plan CHEP07: September 2-7, 2007 Zhenping (Jane) Liu for the BNL RACF Storage Group.
Enabling Grids for E-sciencE EGEE-II INFSO-RI Status of SRB/SRM interface development Fu-Ming Tsai Academia Sinica Grid Computing.
Bologna, March 30, 2006 Riccardo Zappi / Luca Magnoni INFN-CNAF, Bologna.
Open Science Grid Consortium Storage on Open Science Grid Placing, Using and Retrieving Data on OSG Resources Abhishek Singh Rana OSG Users Meeting July.
Riccardo Zappi INFN-CNAF SRM Breakout session. February 28, 2012 Ingredients 1. Basic ingredients (Fabric & Conn. level) 2. (Grid) Middleware ingredients.
EMI INFSO-RI StoRM tutorial Installation and Configuration Riccardo Zappi (INFN-CNAF, Italy) StoRM PT Leader 13 th April 2011, Vilnius, EGI-UF.
EGI-InSPIRE RI EGI-InSPIRE EGI-InSPIRE RI EGI solution for high throughput data analysis Peter Solagna EGI.eu Operations.
EGEE Data Management Services
CASTOR: possible evolution into the LHC era
Databases and DBMSs Todd S. Bacastow January 2005.
Jean-Philippe Baud, IT-GD, CERN November 2007
Scalable sync-and-share service with dCache
StoRM: current status and developments
DPM Installation Configuration
Classic Storage Element
StoRM: a SRM solution for disk based storage systems
Vincenzo Spinoso EGI.eu/INFN
The PaaS Layer in the INDIGO-DataCloud
Chapter 2 Database System Concepts and Architecture
Database System Concepts and Architecture
Data Bridge Solving diverse data access in scientific applications
Open Source distributed document DB for an enterprise
GGF OGSA-WG, Data Use Cases Peter Kunszt Middleware Activity, Data Management Cluster EGEE is a project funded by the European.
Data services on the NGS
Introduction to Data Management in EGI
Distribution and components
T-StoRM: a StoRM testing framework
Study course: “Computing clusters, grids and clouds” Andrey Y. Shevel
CSC 480 Software Engineering
Ákos Frohner EGEE'08 September 2008
Introduction to J2EE Architecture
The INFN Tier-1 Storage Implementation
University of Technology
Data Management cluster summary
A Web-Based Data Grid Chip Watson, Ian Bird, Jie Chen,
Lecture 1: Multi-tier Architecture Overview
Database Environment Transparencies
Introduction to Active Directory Directory Services
INFNGRID Workshop – Bari, Italy, October 2004
Database System Architectures
StoRM disk management middleware
Presentation transcript:

StoRM Architecture and Daemons

Outline SRM standard What is StoRM Deployment Scenario StoRM-Tutorial for supporter, IGI, Bologna, Italy

Storage Resources The storage resource is the combination of the storage system and the controlling software (e.g. file system) Different storage resources offer: different levels of Quality of Service (QoS) different access technologies StoRM-Tutorial for supporter, IGI, Bologna, Italy

Storage Resource Manager (SRM) The Storage Resource Manager (SRM) Interface describes the management service of a Storage Element. SRM services according to the SRM interface hide storage characteristics and allow interoperability. StoRM-Tutorial for supporter, IGI, Bologna, Italy

SRM: key concepts Lifetime of a file File pinning Storage classes volatile with a fixed lifetime, durable or permanent File pinning to ensure a file is not canceled during operation Storage classes to identify different quality of storage resources to enable a separate set Space reservation to ensure the request space is available for the whole life of the application since the beginning StoRM-Tutorial for supporter, IGI, Bologna, Italy

StoRM StoRM is a Storage Resource Manager (SRM) solution designed to leverage the advantages of cluster file systems and standard POSIX file systems in a Grid environment. Current team: Riccardo Zappi, PT leader Elisabetta Ronchieri Michele Dibenedetto StoRM-Tutorial for supporter, IGI, Bologna, Italy

StoRM product factsheet Part of EMI project Sources and binary distributed with EMI releases Developed by INFN Evolving since the late 2005 Used in production in about 50 sites including Italian TIER-1 at CNAF Manages over 20PB di storage space Composed by various sub-components BE, FE, GFTP, GHTTPS, … StoRM-Tutorial for supporter, IGI, Bologna, Italy

StoRM main characteristics Relies on the aggregation functionalities provided by file systems Designed to be independent from the different file system supported Highly scalable and configurable, it can be used at site with different size and requirements Allows to expose in Grid via SRM interface files stored in a standard file system StoRM-Tutorial for supporter, IGI, Bologna, Italy

StoRM and Cluster FS StoRM is designed to take advantage from high performing cluster file system, as GPFS and Lustre, but it supports also every standard POSIX FS (ext3, ext4, xfs, …) . It allows direct access (through the protocol file:// ) to the storage resource, as well as other grid protocol as https, gsiftp and rfio. StoRM-Tutorial for supporter, IGI, Bologna, Italy

StoRM role in a Grid Site StoRM-Tutorial for supporter, IGI, Bologna, Italy

Multi-layer Architecture of StoRM Front-End exposes the web service interface manages user authentication Manages connection with clients Communicates directly with Back-End Stores asynchronous request into the data base Retrieves request status from the data base StoRM-Tutorial for supporter, IGI, Bologna, Italy

Multi-layer Architecture of StoRM Back-End core of StoRM executes all the requests manages user authorization enforces permissions on files and directories interacts with other grid services interacts with file system through a driver mechanism StoRM-Tutorial for supporter, IGI, Bologna, Italy

Stateless components Files properties are stored in Extended Attributes managed by underlying file system. The Database is used to store: SRM request and status Space metadata (e.g. space usage) information FS Namespace is not stored anywhere! Underlying file system provide it very well, so why duplicate it? Space metadata are recreated at service starting, if needed, so: Accidental loss of the DB simply lead to failing ongoing SRM requests.

The other components StoRM-globus-GridFTP Implement the Data Storage Interface (DSI) Seamless integration with Globus GridFTP Calculate Adler32 checksum on the fly Store calculated checksum into an Extended Attribute of the file itself Transparent to the client or remote party.

The other components StoRM-Checksum Calculate the checksum of files already stored on the storage RestFul web service developed as a stand-alone service embedding Jetty container. accept requests containing the file path and the algorithm to be used to compute the checksum elaborates a maximum of concurrent requests depending on how many threads are configured. The result of the elaboration is returned to the client, typically represented by the BE component.

The other components StoRM-GridHTTPs provides http and https protocol as data access protocol able to manage plain requests and requests over SSL providing using a valid x509 certificate Implemented as a Servlet, so it works in any Servlet container Act as a PEP for data access. PDP is currently act by the StoRM-BE.

Deployment scenario Deployment schema: depending on site requirements and load expected. Components replication and distribution: All the components, except for Back-End, are configurable in pool. Notes: Front-End server is CPU intensive due to GSI auth. GridFTP, CheckSum, GridHTTPs servers are CPU and I/O intensive.

Component pooling DNS Aliasing: Internal load balancer: Front-End, GridFTP, CheckSum, GridHTTPs servers. Internal load balancer: Allows a more smart load balancing policies Already in place for GridFTP Coming with next version for GridHTTPs and Checksum.

Simplest deployment All in one host Suggested for small SE One single Front-End One single Back-End One single GridFTP MySQL server Suggested for small SE Low transfer traffic not means with small space size.. Few requests for second

Simplest deployment Scaling the number of FE on the load expected Suggested for: Supporting burst of requests 1 FE = 400 req/min 2 FE = 700 req/min Supporting for fault tolerance DNS alias + Nagios monitoring

All the StoRM components

Most common deployment

Why scaling FE?