Øystein Torbjørnsen

Evaluating the effectiveness of fault tolerance in replicated database management systems

Database management systems (DBMS) achieve high availability and fault tolerance usually by replication. However fault tolerance does not come for free. Therefore, DBMSs serving critical applications with real time requirements must find a trade of between fault tolerance cost and performance. The purpose of this study is two-fold. It evaluates the effectiveness of DBMS fault tolerance in the presence of corruption in database buffer cache, which poses serious threat to the integrity requirement of the DBMSs. The first experiment of this study evaluates the effectiveness of fault tolerance, and the fault impact on database integrity, performance, and availability on a replicated DBMS, ClustRa, in the presence of software faults that corrupt the volatile data buffer cache. The second experiment identifies the weak data structure components in the data buffer cache that give fatal consequences when corrupted, and suggest the need for some forms of guarding them individually or collectively.

Inverted indexes vs. bitmap indexes in decision support systems

Bitmap indexes are widely used in Decision Support Systems (DSSs) to improve query performance. In this paper, we evaluate the use of compressed inverted indexes with adapted query processing strategies from Information Retrieval as an alternative. In a thorough experimental evaluation on both synthetic data and data from the Star Schema Benchmark, we show that inverted indexes are more compact than bitmap indexes in almost all cases. This compactness combined with efficient query processing strategies results in inverted indexes outperforming bitmap indexes for most queries, often significantly.

XLeaf: Twig Evaluation with Skipping Loop Joins and Virtual Nodes

XML indexing and search has become an important topic, and twig joins are key building blocks in XML search systems. This paper describes a novel approach using a nested loop twig join algorithm, which combines several existing techniques to speed up evaluation of XML queries. We combine structural summaries, path indexing and prefix path partitioning to reduce the amount of data read by the join. This effect is amplified by only reading data for leaf query nodes, and inferring data for internal nodes from the structural summary. Skipping is used to speed up merges where query leaves have differing selectivity. Multiple access methods are implemented as materialized views instead of succinct secondary indexes for better locality. This redundancy is made affordable in terms of space by using compression in a back-end with columnar storage. We have implemented an experimental prototype, which shows a speedup of two orders of magnitude on XPath queries with value predicates, when compared to existing open source and commercial systems using a subset of the techniques. Space usage is also improved.

Designing an ultra highly available DBMS (tutorial session)

Most database management systems available today are systems designed for general use. Certainly, some compromises have been done to satisfy the the most common users and the largest markets. One application which has been mostly ignored until now is the network equipment made for the telco operators. The equipment used in the telco industry has requirements which have been quite different from the typical database applications, especially with respect to availability and real-time performance. This has caused the telco manufacturers to develop their own hardware, operating systems, programming languages and database management systems. There has been little standardization and cooperation causing a wide variety of solutions both between the vendors and within the vendors themselves. This has been possible because of high prices, long product development cycles and long product lifetimes. Due to the deregulation of the market, increased competition, better price, quality and performance of standard products and an emerging Internet market with telco requirements, standard SW and HW products are becoming more and more fit for the telco market. Any database system trying to replace the proprietary solutions must satisfy the basic telco requirements: availability (> 99.999%);

Cost of ensuring safety in distributed database management systems

Generally, applications employing database management systems (DBMS) require that the integrity of the data stored in the database be preserved during normal operation as well as after crash recovery. Preserving database integrity and availability needs extra safety measures in the form of consistency checks. Increased safety measures inflict adverse effect on performance by reducing throughput and increasing response time. This may not be agreeable for some critical applications and thus, a tradeoff is needed. This study evaluates the cost of extra consistency checks introduced in the data buffer cache in order to preserve the database integrity, in terms of performance loss. In addition, it evaluates the improvement in error coverage and fault tolerance, and occurrence of double failures causing long unavailability, with the help of fault injection. The evaluation is performed on a replicated DBMS, ClustRa. The results show that the checksum overhead in a DBMS inflicted with a very high TPC-B-like workload caused a reduction in throughput up to 5%. The error detection coverage improved from 62% to 92%. Fault injection experiments shows that corruption in database image went down from 13% to 0%. This indicates that the applications that require high safety, but can afford up to 5% performance loss can adopt checksum mechanisms.