{"id":105025,"date":"2022-09-05T21:36:50","date_gmt":"2022-09-05T19:36:50","guid":{"rendered":"https:\/\/prohoster.info\/blog\/novosti-interneta\/predstavlena-raspredelyonnaya-operaczionnaya-sistema-dbos-rabotayushhaya-poverh-subd"},"modified":"2022-09-05T21:36:50","modified_gmt":"2022-09-05T19:36:50","slug":"predstavlena-raspredelyonnaya-operaczionnaya-sistema-dbos-rabotayushhaya-poverh-subd","status":"publish","type":"post","link":"https:\/\/prohoster.info\/en\/blog\/news\/predstavlena-raspredelyonnaya-operaczionnaya-sistema-dbos-rabotayushhaya-poverh-subd","title":{"rendered":"The distributed operating system DBOS has been introduced, functioning on top of the DBMS.","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"
The DBOS (DBMS-oriented Operating System) project has been introduced, developing a new operating system for executing scalable distributed applications. A key feature of the project is the use of a DBMS for storing applications and system state, with access to the state organized exclusively through transactions. The project is being developed by researchers from the Massachusetts Institute of Technology, the University of Wisconsin, Stanford University, Carnegie Mellon University, and companies like Google and VMware. The developments are released under the MIT license. <\/p>\n
Components for hardware interaction and low-level memory management services have been moved to a microkernel. The capabilities provided by the microkernel are utilized to launch a layer with the DBMS. High-level system services that ensure application execution interact only with the distributed DBMS and are separated from the microkernel and components specific to any given system. <\/p>\n
Building on top of a distributed DBMS allows system services to be inherently distributed and not tied to a specific node, which differentiates DBOS from traditional cluster systems where each node runs its own instance of the operating system, above which separate cluster schedulers, distributed file systems, and network managers operate.
<\/center> <\/p>\n
It is noted that using modern distributed DBMS, which store data in memory and support transactions, such as VoltDB and FoundationDB, as the foundation for DBOS can provide the performance necessary for efficiently executing many system services. The DBMS can also host data for the scheduler, file system, and IPC. These DBMS scale excellently, ensuring transaction atomicity and isolation, can handle petabytes of data, and provide tools for managing access and tracking data flows. <\/p>\n
Among the advantages of the proposed architecture is the significant enhancement of analytics capabilities and reduced code complexity through the use of conventional database queries in services of the operating system, where transaction implementation and means for ensuring high availability are offloaded to the database management system (this functionality can be implemented on the DBMS side once and utilized by the OS and applications). <\/p>\n
For instance, a cluster scheduler can store information about tasks and handlers in database tables and execute scheduling operations as regular transactions, blending imperative code and SQL. The use of transactions simplifies issues such as multithreading management and fault recovery, as transactions ensure consistency and resilience of the state. In the context of the scheduler example, transactions allow for simultaneous access to shared data and guarantee state integrity in case of failures. <\/p>\n
The mechanisms provided by the DBMS for logging and data analysis can be utilized to track access and changes in application states, monitor, debug, and maintain security. For example, after detecting unauthorized access to the system, it is possible to determine the scale of the breach by executing SQL queries to identify all operations performed by processes that accessed sensitive information. <\/p>\n
The project has been in development for over a year and is currently in the prototyping stage of individual components of the architecture. At present, a prototype of operational services built on top of the DBMS has been prepared, such as file systems, IPC, and a scheduler, as well as a programming environment that provides an interface for running applications based on the FaaS (function-as-a-service) model. <\/p>\n
In the next stage of development, a complete software stack for distributed applications is planned to be provided. Currently, VoltDB is being used as the database management system in experiments, but discussions are underway to create a custom data storage layer or to implement missing functionalities in existing DBMS. The question of which components should operate at the core level and which can be implemented on top of the DBMS is also under discussion. \n Source: opennet.ru<\/a> <\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"
![\"The](\"\/wp-content\/uploads\/2022\/09\/3f504305269ef739afa07f5b2d737df7.jpg\")
<\/center> <\/p>\n