On January 28, 2016 in the Moscow office of Mail.Ru Group there will take place the second meeting of Tarantool meetup. If someone else does not know: Tarantool is the NoSQL In-Memory DBMS open source which is created for ensuring the greatest possible performance. On the second mitap we will consider the main benefits and features of Tarantool, we will tell about the experience of use of this product and plans for the future. First of all this meeting will be interesting to the developers, Unix-system administrators and other specialists anyway working with databases. You watch the program of a meeting under a cat.
18:30 — collecting of participants.
19:00 — "A tarantula — as we came to development of own DBMS", Denis Anikin, Mail.Ru Group
I will tell about the most widespread problems of the high-loaded projects, about their reasons and standard methods of their solution. You learn why such projects at some point cease to work and as in this case Tarantool can help.
The solution on development of own DBMS (database management system) of general purpose usually ripens long. Before rushing to this abyss, we at first look for whether it is possible to solve the current problems the existing tools, including creating over them or integrating into different solutions. And only when before us the light image of the ideal database possessing all necessary features appears we rush to all heavy and we begin to create.
Quite so there was also Tarantool — one of the fastest databases in the world which is widely applied in Mail.Ru Group and beyond its limits. Tarantool was from the very beginning developed by separate command, in open source codes and with a view to the general use, and as a result we received an excellent product — DBMS of general purpose with a unique set of features.
19:40 — "The user sessions in Badoo", Dmitry Novikov, Badoo
Before we began to use Tarantool, the user sessions in Badoo were stored in the samopisny demon. The demon was good, but with him there were some problems. For example, it stored data only in memory. And in general everything that happened to sessions, began to give a certain inconvenience. Therefore that software development brought joy and pleasure, we decided to remake cardinally the system working with sessions.
20:20 — "Practice of use of Tarantool in products: experience of Wallarm", Alexander Golovko, Wallarm
Practical results of two years' experience of use of Tarantool in alienable software solution will be presented in the report. We needed to perform asynchronous analytics according to short-lived objects of NTTR-requests. We studied a number of possible candidate solutions of this task, and as a result selected Tarantool and the concept of application-server. In the report comparison of characteristics of all considered means will be carried out and arguments and criteria at the choice of architecture are adduced. The final part of performance is devoted to problems which developers faced at implementation, operation and migration on new Tarantool versions.
At design of the system of protection of web applications, we faced need to correlate data on several already processed HTTP requests. The most effective solution of this task was separation of logic of request processing between synchronous computation process of parameters for the analysis and asynchronous process of the analytics. Actually, we took out in Tarantool all post-processing of NTTR-requests after their primary analytics on the party of Nginx. It became as a result possible to find different behavioural attacks to web applications, for example, search of passwords or codes.
21:05 — end of a meeting.
Registration is already open!
This article is a translation of the original post at habrahabr.ru/post/273845/
If you have any questions regarding the material covered in the article above, please, contact the original author of the post.
If you have any complaints about this article or you want this article to be deleted, please, drop an email here: firstname.lastname@example.org.
We believe that the knowledge, which is available at the most popular Russian IT blog habrahabr.ru, should be accessed by everyone, even though it is poorly translated.
Shared knowledge makes the world better.