Back to Search
Start Over
Developing a framework architecture of a secure big data lake
- Publication Year :
- 2022
- Publisher :
- СанкÑ-ÐеÑеÑбÑÑгÑкий полиÑÐµÑ Ð½Ð¸ÑеÑкий ÑнивеÑÑиÑÐµÑ ÐеÑÑа Ðеликого, 2022.
-
Abstract
- Рданной вÑпÑÑкной квалиÑикаÑионной ÑабоÑе пÑедÑÑавлена ÑазÑабоÑка аÑÑ Ð¸ÑекÑÑÑÑ ÑÑеймвоÑка заÑиÑÑнного озеÑа болÑÑÐ¸Ñ Ð´Ð°Ð½Ð½ÑÑ . Ðа оÑÐ½Ð¾Ð²Ñ Ð²Ð·ÑÑ Data Lake Architecture Framework (DLAF), ÑазÑабоÑаннÑй иÑÑледоваÑелÑми даннÑÑ Ð¨ÑÑÑгаÑдÑкого УнивеÑÑиÑеÑа. Ð ÑабоÑе иÑÑледована меÑÐ¾Ð´Ð¾Ð»Ð¾Ð³Ð¸Ñ DLAF, его аÑпекÑÑ Ð¸ из взаимодейÑÑвиÑ. РазÑабоÑка ÑÑедÑÑва обеÑпеÑÐµÐ½Ð¸Ñ Ð¸Ð½ÑоÑмаÑионной безопаÑноÑÑи озеÑа даннÑÑ Ð½Ð° аÑÑ Ð¸ÑекÑÑÑном ÑÑовне поÑÑебовала ÑÑаÑелÑного оÑбоÑа ÑÑенаÑиев аÑак и моделей безопаÑноÑÑи. ÐÑÐ±Ð¾Ñ Ð¿ÑоводилÑÑ Ñ ÑÑÑÑом ÑÐ°ÐºÐ¸Ñ Ð³Ð»Ð°Ð²Ð½ÑÑ Ð¾ÑлиÑиÑелÑнÑÑ ÑеÑÑ Ð¾Ð·ÐµÑа даннÑÑ , как возможноÑÑи загÑÑзки, обÑабоÑки и вÑгÑÑзки даннÑÑ Ð°Ð±ÑолÑÑно лÑбого Ñипа, а Ñакже взаимодейÑÑвие Ñ Ð»ÑбÑм иÑÑоÑником даннÑÑ . ÐÑбÑаннÑе модели безопаÑноÑÑи в Ñвоей оÑнове ÑоÑÑоÑÑ Ð¸Ð· ÑолевÑÑ Ð¸Ð»Ð¸ аÑÑибÑÑнÑÑ Ð¿Ð¾Ð»Ð¸Ñик. ÐаннÑе модели пÑеÑеÑпели модиÑикаÑии под нÑÐ¶Ð´Ñ Ð¾Ð·ÐµÑа даннÑÑ . ÐбеÑпеÑение заÑиÑÑ Ð´Ð°Ð½Ð½ÑÑ Ð²Ð¾Ð·Ð»Ð¾Ð¶ÐµÐ½Ð¾ на два компоненÑа: Global Monitoring Tool и Check Sum Controller. ÐаннÑе компоненÑÑ Ð²ÐºÐ»ÑÑÐµÐ½Ñ Ð² полиÑики модели безопаÑноÑÑи на ÑÑапе пÑоекÑиÑованиÑ. ÐеÑвÑй ÐºÐ¾Ð¼Ð¿Ð¾Ð½ÐµÐ½Ñ Ñеализован в виде композиÑии моделей маÑинного обÑÑÐµÐ½Ð¸Ñ Ð¸ нÑждаеÑÑÑ Ð² пÑедваÑиÑелÑном обÑÑении. Ðн оÑвеÑÐ°ÐµÑ Ð·Ð° мониÑоÑинг вÑÐµÑ Ð¾Ð¿ÐµÑаÑий озеÑа даннÑÑ Ð¸ Ð¸Ñ ÐºÐ»Ð°ÑÑиÑикаÑии на безопаÑнÑе и вÑедоноÑнÑе. ÐÑоÑой ÐºÐ¾Ð¼Ð¿Ð¾Ð½ÐµÐ½Ñ Ð¾Ð±ÑабаÑÑÐ²Ð°ÐµÑ Ð»Ñбое взаимодейÑÑвие Ñ ÐºÐ¾Ð½ÑÑолÑной ÑÑммой Ñайлового обÑекÑа в ÑеÑение вÑего пÑебÑÐ²Ð°Ð½Ð¸Ñ Ð² озеÑе даннÑÑ . ÐÑиведÑннÑй в ÑабоÑе инÑÑÑÑменÑалÑнÑй аÑÐ´Ð¸Ñ Ð¸Ð½ÑоÑмаÑионной безопаÑноÑÑи дал обÑекÑивнÑе оÑенки ÑазÑабоÑанного ÑеÑÐµÐ½Ð¸Ñ Ð¸ ÑекомендаÑии по вÑбоÑÑ Ð¼Ð¾Ð´ÐµÐ»ÐµÐ¹ безопаÑноÑÑи.<br />This final qualification paper presents the development of a protected big data lake framework architecture. It is based on the Data Lake Architecture Framework (DLAF) developed by data scientists at the University of Stuttgart. The paper investigates the DLAF methodology, its aspects and from interactions. Developing a data lake information security tool at the architectural level required a careful selection of attack scenarios and security models. The selection was made taking into account the main features of the data lake, such as the ability to download, process and upload data of absolutely any type, as well as interaction with any data source. The selected security models consist fundamentally of role or attribute policies. These models have undergone modifications for the needs of the data lake. Data protection is assigned to two components: Global Monitoring Tool and Check Sum Controller. These components are included in the security model policies at the design stage. The first component is implemented as a composition of machine learning models and needs to be pre-trained. It is responsible for monitoring all data lake operations and classifying them into safe and malicious ones. The second component handles any interaction with the checksum of the file object during the entire stay in the data lake. The information security instrumental audit presented in this paper provided objective evaluations of the developed solution and recommendations for security model selection.
Details
- Language :
- Russian
- Database :
- OpenAIRE
- Accession number :
- edsair.doi...........4a3b1ea173f2a93395c043fc03baf9ac
- Full Text :
- https://doi.org/10.18720/spbpu/3/2023/vr/vr23-508