Event Agenda and Speaker info
10:45 - Registration
11:00 - Tea
11:15 - Keynote: Challenges in scaling cloud storage - Srivibhavan (Vibhav) Balaram, Founder of CloudeByte
Challenges in cloud (shared storage) are many and traditional storage architectures were not geared to solve them. This keynote will highlight the “noisy neighbour” issue and the issues of supporting performance sensitive applications in a cloud storage environment.
Vibhav is a seasoned General Manager and Entrepreneur with an exceptional technology background. In his long career he worked for the NIC (GoI), TCS, Novell, HP, Citrix, NetApp, SAP, Oracle and EMC. In his early career he was engaged in building code for PC to mainframe communication, SNA, TCP/IP and OSI stacks. Later he built network management software with SNMP. He started his entrepreneurial journey with CloudByte – a storage start up. He holds a bachelor degree in Electronics and Communication from Mysore University and short sessions with IIT Delhi for a doctoral work and later in IIM Bangalore and ISB Hyderabad
11:30 - Performance Characterization in Large distributed file system with GlusterFS - Neependra Khare, Principal Software Engineer, Red Hat
In the era of cloud computing, distributed file-systems have become the backbone of cloud infrastructure. After choosing the right file-system it becomes very important to get best possible performance for different workloads as we scale. In the talk we'll understand how to characterize the performance of GlusterFS for given workload/application. How application characteristics effect the performance in distributed file-system environment ? We'll see how to quickly identify and isolate performance bottlenecks and learn about the best practices and tuning options available for GlusterFS.
Neependra has ~9 years of experience in IT industry. He has worked as SysAdmin, support engineer, file system developer. Currently he is working with Red Hat as Performance Engineer and looking after performance for Red Hat Storage, which enterprise version of GlusterFS
12:15 - Scaling using event driven programming with Perl, A tutorial - Aveek Mishra, Principal Engineer, Yahoo
Building highly concurrent systems, such as large-scale Internet services, requires managing many information flows at once and maintaining peak throughput when demand exceeds resource availability. In addition, any platform supporting Internet services must provide high availability and be able to cope with burstiness of load. Many approaches to building concurrent systems have been proposed, which generally fall into the two categories of threaded and event driven programming. In this talk, I will primarily dwell on event driven programming for back end systems and specifically how Perl AnyEvent framework can be used to build highly concurrent and scalable servers.
1:00 - Lunch
2:00 - Quality of Service in Cloud Systems - Dr. J. Lakshmi, Principal Research Scientist, Supercomputer Education and Research Center, IISc Bangalore
Clouds are emerging as computing platforms that use dynamic configuration of software and hardware. The motivation for adopting clouds is being able to use a service without owning it, and paying by usage rather than for ownership and maintenance. Enabling architectures for clouds are service oriented architectures including both the software and hardware. Virtualization technologies give impetus to this kind of usage since they seamlessly support on-demand enabling of services with the required applications and their requisite platforms and hardware. Cloud models support multi-tenancy of applications so as to reuse the hardware and/or software across multiple clients. Since clouds allow for multi-tenancy of applications over the same hardware, guaranteeing application performance will govern, to a large extent, the usage models of clouds. As clouds gain greater acceptance, user expectations from the clouds will start moving from best-effort service to guaranteed service. Hence quality of service (QoS) will be a dominant consideration for cloud adaptation. QoS has many facets and will depend on the aspect that is crucial for the user. Application specific performance, as response time or throughput, application security varying from data integration and consistency to privacy, service availability, etc., are some of the QoS considerations that clouds need to address. In this talk I will share some findings on the challenges in assuring performance on virtualized platforms specifically for I/O intensive workloads.
Lakshmi is with SERC, IISc since 1997 managing their Computational center supporting HPC scientific applications. Earlier Lakshmi has worked in ADA in the same role. She also heads the Cloud Systems Lab of SERC. Her research interests lie in operating systems particularly for system virtualization and distributed systems. Currently she is working in exploring QoS to support performance, security, availability, etc.
2:45 - Serving Display Advertisement at Scale - Varun Verma, Principal Engineer and Arvind Kota, Senior Engineering Manager, Yahoo
In this session Arvind and Varun would let us look into how Yahoo! handles it's infrastructure to serve billions of display advertisement everyday
Arvind has been with Yahoo! for nine years and currently manages the Guaranteed Display Serving Engineering team in Bangalore. He is an avid runner, with a penchant for marathons. Varun is a 'newbie' to Yahoo!, clocking in just over 3 years and is currently the Service Engineering Architect For Display Advertising in Bangalore.
3:30 - Open forum
3:45 onwards - Tea, snacks and networking.
Enter into the EGL park from Koramangala Indiranagar intermediate ring road. The first building on your left is the venue, Yahoo!. This is the Sunriver building. Take the first left to get into the parking lot. Ask the security for help and tell them that you have come for the LSPE event at Yahoo and they will help you park.