Peer-to-peer systems have recently gained a lot of attention in the academic community especially through the design of KBR (Key-Based Routing) algorithms and DHT (Distributed Hash Table)s. On top of these constructs were built promising applications such as video streaming applications but also storage infrastructures benefiting from the availability and resilience of such scalable network protocols. Unfortunately, rare are the storage systems designed to be scalable and fault-tolerant to Byzantine behaviour, conditions required for such systems to be deployed in an environment such as the Internet. Furthermore, although some means of access control are often provided, such file systems fail to offer the end-users the flexibility required in order to easily manage the permissions granted to potentially hundreds or thousands of end-users. In addition, as for centralised file systems which rely on a special user, referred to as root on Unices, distributed file systems equally require some tasks to operate at the system level. The decentralised nature of these systems renders impossible the use of a single authoritative entity for performing such tasks since implicitly granting her superprivileges, unacceptable configuration for such decentralised systems. This thesis addresses both issues by providing the file system objects a completely decentralised access control and administration scheme enabling users to express access control rules in a flexible way but also to request administrative tasks without the need for a superuser. A prototype has been developed and evaluated, proving feasible the deployment of such a decentralised file system in large-scale and untrustworthy environments.

Nowadays there is wide variety of programming languages and platforms providing com-plex set of tools for software development -- from mobile and desktop applications up to server applications. Java is one of the most popular choices for software project develop-ment since many years. Many changes and improvements of the Java language occurred throughout 23 years of development of the platform. Current version brings many changes description of whom is considered to be beneficial not only for developers but also for the community around Java platform. This thesis covers the changes that were introduced in the release of Java 7. The thesis lists the short review of main historical events that happened since the beginning of the platform and analyses changes in libraries as well as describes syntactical changes. Primary focus is placed on the NIO (New I/O) library, whose current name is Non-Blocking I/O (NIO.2). Even though massive changes had occurred in this library, not many people actually noticed anything. This unfortunate situation is now described as a quiet revolution. Last but not least a description of changes in other libraries and Java Virtual Machine is presented.

Btrfs je copy-on-write linuxový souborový systém, který obsahuje vlastního správce oddílů a podporuje efektivní operace se snapshoty. Online deduplikace dat je metoda odstranění duplicitních bloků dat ještě předtím než jsou zapsány na disk. Tato funkcionalita šetří místo na uložišti avšak může v jistých případech znamenat pokles výkonu při zápisu. Hlavním využitím jsou soubory obrazů disků virtuálních strojů. Takové soubory obsahují z velké části stejná data (stejný operační systém) a tedy tyto části lze zapsat na disk pouze jednou. Tato diplomová práce se zabývá návrhem a implementací online deduplikace pro souborový systém Btrfs.

Most of the software stack is built upon C today. C is a very flexible language, but the flexibility also brings some safety risks, particularly when handling memory through pointers. Rust is a new programming language which can guarantee memory safety without performance-heavy runtime services such as garbage collection. In this article, two partial file systems are implemented based on the design of EXT2. One system is implemented in C — Nafs — and one system is implemented in Rust — Rufs. A number of benchmarks are also developed, with the purpose of testing the most common features of a file system. After running the benchmarks, the results showed that Nafs provided better performance for all but one feature. There could be many reasons for this, and some hypotheses are discussed. Aspects like different compiler optimizations, the use of pointer dereferencing vs. data structure representation, using dynamic memory, and using system calls are considered. Some optimizations to Rufs are also implemented, and their impact analyzed. Earlier research has shown that Rust can guarantee memory safety while still providing good performance. It has also shown that Rust can be used to implement system programs, such as operating system kernels. Over the course of this article, it is shown that safe Rust can be used to implement a file system — thereby guaranteeing a memory safe program. It is also shown that a file system implemented in safe Rust provides worse performance than a similar system written in C. Future work will have to tell whether a file system can be as performant in Rust as in C, by using different implementation methods or unsafe Rust for some of the most performance critical parts of the system.

Les systèmes de fichiers géo-distribués souffrent de latences élevées et de partitions réseau. À cause de cela, et pour assurer une haute disponibilité, de tels systèmes effectuent généralement des mises à jour localement, sans latence, et les propagent ensuite en arrière-plan. Cette réplication optimiste est confrontée à deux défis majeurs : (i) détecter les conflits entre les mises à jour simultanées et les résoudre d'une manière significative pour les utilisateurs, tout en maintenant les invariants d'intégrité du système; et (ii) la prise en charge d'applications qui n'ont pas été conçues pour gérer les anomalies de concurrence. Les systèmes de fichiers géo-distribués optimistes existants ne permettent pas de relever ces défis. Par exemple, Dropbox ne supporte pas les liens matériels. Le système de fichiers AndrewFS échoue sur certains changements de noms de répertoires; et tous les systèmes existants utilisent la résolution automatique des conflits qui viole la sémantique POSIX. Nous présentons notre solution aux problèmes posés ci-dessus dans la conception et la mise en œuvre d'un prototype de système de fichiers géo-distribué, nommé Tofu. Sa conception inclut une nouvelle abstraction de session pour prendre en charge l'API, tout en permettant des mises à jour optimistes. Il est capable de détecter tous les conflits sur ces structures de données et de les résoudre d'une façon que nous pensons que les utilisateurs trouveront raisonnable. Les expériences montrent que Tofu est hautement évolutif et qu'il entraîne des surcoûts linéaires, améliorant ainsi les systèmes académiques et industriels existants. / Geo-distributed file systems suffer from high latency and network partitions. Because of this, and to ensure high availability, such systems typically commit updates locally, with no latency, and propagate them in the background. Such optimistic replication faces two major challenges: (i) detecting conflicts between concurrent updates and resolving them in a way meaningful for users, while maintaining system integrity invariants; and (ii) supporting legacy applications that are not prepared to deal with concurrency anomalies. Existing optimistic geo-distributed file systems fall short of addressing the challenges. For instance, Dropbox does not support hard links; Andrew File System fails on some concurrent renaming of directories; and all existing systems use automatic conflict resolution that violates the legacy POSIX semantics. We present our solution to the above problems in the design and implementation of a prototype geo-distributed file system, named Tofu. Its design includes a new session abstraction to support the legacy API, while allowing optimistic updates. Unlike previous approaches, our solution is based on a formal model covering all aspects of a Unix-like file system, including directories, inodes, hard links, etc. It is able to detect all conflicts on those data structures, and resolves them in a way that we believe users will find generally reasonable. Experiments show that Tofu is highly scalable, and incurs linear overhead, improving over existing academic and industrial systems.

Blockchain is an interesting technology, with different projects developing every day since it first gained its light back in 2008. More and more research finds blockchain useful in several different sectors. One of the sectors being healthcare, specifically for electronic health records (EHR). EHR contains highly sensitive data which is critical to protect and, just in the year 2019, 41,232,527 records were deemed stolen. Blockchain can provide several benefits when it comes to EHR, such as increased security, availability, and privacy, however, it needs to be done correctly. Due to blockchain being a rather novel technology, there is room for improvement when it comes to integrating blockchain with EHR. In this thesis a framework for EHR in the healthcare sector is proposed, using Ethereum based smart contracts together with decentralized off-chain storage using InterPlanetary File System (IPFS) and strong symmetric encryption. The framework secures the records and provides a scalable solution. Furthermore, a discussion and evaluation regarding several security aspects that the framework excels on as well as what the framework could improve on.

The relevance of data processing and analysis today cannot be overstated. The convergence of several technological advancements has fostered the proliferation of systems and infrastructure that together support the generation, transmission, and storage of nearly 15,000 exabytes of digital, analyzabledata. The Hadoop Distributed File System (HDFS) is an open source system designed to leverage the storage capacity of thousands of servers, and is the file system component of an entire ecosystem of tools to transform and analyze massive data sets. While HDFS is used by organizations of all sizes, smaller ones are not as well-suited to organically grow their clusters to accommodate their ever-expanding data sets and processing needs. This is because larger clusters are concomitant with higher investment in servers, greater rates of failures to recover from, and the need to allocate moreresources in maintenance and administration tasks. This poses a potential limitation down the road for organizations, and it might even deter some from venturing into the data world altogether. This thesis addresses this matter by presenting a novel implementation of HopsFS, an already improved version of HDFS, that requires no user-managed data servers. Instead, it relies on S3, a leading object storage service, for all its user-data storage needs. We compared the performance of both S3-based and regular clusters and found that such architecture is not only feasible, but also perfectly viable in terms of read and write throughputs, in some cases even outperforming its original counterpart. Furthermore, our solution provides first-class elasticity, reliability, and availability, all while being remarkably more affordable. / Relevansen av databehandling och analys idag kan inte överdrivas. Konvergensen av flera tekniska framsteg har främjat spridningen av system och infrastruk-tur som tillsammans stöder generering, överföring och lagring av nästan 15,000 exabyte digitala, analyserbara data. Hadoop Distributed File System (HDFS) är ett öppen källkodssystem som är utformat för att utnyttja lagringskapaciteten hos tusentals servrar och är filsystemkomponenten i ett helt ekosystem av verktyg för att omvandla och analysera massiva datamängder. HDFS används av organisationer i alla storlekar, men mindre är inte lika lämpade för att organiskt växa sina kluster för att tillgodose deras ständigt växande datamängder och behandlingsbehov. Detta beror på att större kluster är samtidigt med högre investeringar i servrar, större misslyckanden att återhämta sig från och behovet av att avsätta mer resurser i underhålls- och administrationsuppgifter. Detta utgör en potentiell begränsning på vägen för organisationer, och det kan till och med avskräcka en del från att våga sig helt in i datavärlden. Denna avhandling behandlar denna fråga genom att presentera en ny implementering av HopsFS, en redan förbättrad version av HDFS, som inte kräver några användarhanterade dataservrar. Istället förlitar sig det på S3, en ledande objektlagringstjänst, för alla dess användardata lagringsbehov. Vi jämförde prestandan för både S3-baserade och vanliga kluster och fann att sådan arkitektur inte bara är möjlig, utan också helt livskraftig när det gäller läs- och skrivgenomströmningar, i vissa fall till och med bättre än dess ursprungliga motsvarighet. Dessutom ger vår lösning förstklassig elasticitet, tillförlitlighet och tillgänglighet, samtidigt som den är anmärkningsvärt billigare.

No description available.

No description available.

No description available.