In-Memory Computing Summit 2017 San Francisco
Not so long ago, at the end of October, another In-Memory Computing Summit was held in Silicon Valley, of which GridGain became one of the organizers and sponsors.
Now you can get free access to video materials and slides, which are divided into two sections: the main theses and reports .
')
More than 40 reports on distributed systems and in-memory calculations were more read at the summit. The speakers included employees from GridGain, Percona, AMY Robotics, Hazelcast, VoltDB Inc., Dell and others.
Below we will separately mention several selected reports.
The representative of Wellington Management talks about the applicability of Apache Ignite in the investment business and the difficulties that may arise during the implementation of the IBOR system.
Distributed platforms such as Apache Ignite make extensive use of horizontal scalability. The more machines in the cluster, the better the performance of the application. Will performance double after adding a second machine to the cluster? Will it increase tenfold when the cluster grows tenfold? Is it always like this? What is the platform responsible for? What are her limitations? And where does what the platform can do ends and the area of responsibility of a qualified engineer who implements it begins?
The path to combining memory and storage is a combination of NVM, the corresponding programming model and NVDIMM. The ability to have fast storage with random access and non-volatility opens up many perspectives. The report talks about the programming model NVM, NVDIMM and the speed advantages that the corresponding equipment gives.
Non-volatile random access memory will allow developers to conduct data operations on multi-terabyte arrays. The report focuses on the characteristics of non-volatile memory and the features that a developer must take into account who wants to work with such memory using Java. Native Linux and Windows APIs, as well as suggestions for adapting Java Collections, are considered.
The report explains how in-memory distributed computing systems can be applied architecturally in financial organizations. Dmitry talks about relevant Apache Ignite features such as ACID, SQL support, persistent storage, replication, security tools, fault tolerance, etc. It also tells about one of the largest such installations in Sberbank, considered 5 major difficulties faced by the bank.
Stay tuned and come join!
Now you can get free access to video materials and slides, which are divided into two sections: the main theses and reports .
')
More than 40 reports on distributed systems and in-memory calculations were more read at the summit. The speakers included employees from GridGain, Percona, AMY Robotics, Hazelcast, VoltDB Inc., Dell and others.
Below we will separately mention several selected reports.
Abe Kleinfeld (GridGain CEO) Keynote
Implementation of Investment Book of Record (IBOR) using Apache Ignite / GridGain (Rafique Awan, Wellington Management)
The representative of Wellington Management talks about the applicability of Apache Ignite in the investment business and the difficulties that may arise during the implementation of the IBOR system.
Scale Out and Conquer: Architectural Decisions Behind Distributed In-Memory Systems (Jacob Zhdanov, GridGain)
linkDistributed platforms such as Apache Ignite make extensive use of horizontal scalability. The more machines in the cluster, the better the performance of the application. Will performance double after adding a second machine to the cluster? Will it increase tenfold when the cluster grows tenfold? Is it always like this? What is the platform responsible for? What are her limitations? And where does what the platform can do ends and the area of responsibility of a qualified engineer who implements it begins?
Realizing the NVM Programming Model and NVDIMMs (Jim Pappas, SNIA)
linkThe path to combining memory and storage is a combination of NVM, the corresponding programming model and NVDIMM. The ability to have fast storage with random access and non-volatility opens up many perspectives. The report talks about the programming model NVM, NVDIMM and the speed advantages that the corresponding equipment gives.
In-Persistent-Memory Computing with Java (Eric Kaczmarek, Intel)
linkNon-volatile random access memory will allow developers to conduct data operations on multi-terabyte arrays. The report focuses on the characteristics of non-volatile memory and the features that a developer must take into account who wants to work with such memory using Java. Native Linux and Windows APIs, as well as suggestions for adapting Java Collections, are considered.
Implementing Durable Memory-Centric Architectures in Large Financial Institutions (Dmitry Setrakyan, GridGain)
linkThe report explains how in-memory distributed computing systems can be applied architecturally in financial organizations. Dmitry talks about relevant Apache Ignite features such as ACID, SQL support, persistent storage, replication, security tools, fault tolerance, etc. It also tells about one of the largest such installations in Sberbank, considered 5 major difficulties faced by the bank.
In-Memory Computing Summit 2018
Next year there will be two more international summits in which you can take part as a speaker or participant. All information will be published on the official website .Stay tuned and come join!
Source: https://habr.com/ru/post/343556/
All Articles