No abstract

Business growth and technology advancements have resulted in growing amounts of enterprise data. To gain valuable business insight and competitive advantage, businesses demand the capability of performing real-time analytics on such data. This, however, involves expensive query operations that are very time consuming on traditional CPUs. Additionally, in traditional database management systems (DBMS), the CPU resources are dedicated to mission-critical transactional workloads. Offloading expensive analytics query operations to a co-processor can allow efficient execution of analytics workloads in parallel with transactional workloads.In this paper, we present a Field Programmable Gate Array (FPGA) based acceleration engine for database operations in analytics queries. The proposed solution provides a mechanism for a DBMS to seamlessly harness the FPGA compute power without requiring any changes in the application or the existing data layout. Using a software-programmed query control block, the accelerator can be tailored to execute different queries without reconfiguration. Our prototype is implemented in a PCIe-attached FPGA system and is integrated into a commercial DBMS platform. The results demonstrate up to 94% CPU savings on real customer data compared to the baseline software cost with up to an order of magnitude speedup in the offloaded computations and up to 6.2x improvement in end-to-end performance.

No abstract

No abstract

We consider a system of identical parallel queues served by a single server and distinguished only by the price charged at entry. A Poisson stream of customers joins the queue by a greedy policy that minimizes a 'disutility' that combines price and congestion. A special case of linear disutility is analyzed for which it is shown that the individually optimal greedy queue join policy is nearly socially optimal. For this queueing system, a Markov decision theoretic framework is formulated for dynamic pricing in the general case. This queueing system has application in the pricing of Internet services. 0-7803-7476-2/02/$17.00 (c) 2002 IEEE. 0-7803-7476-2/02/$17.00 (c) 2002 IEEE.

Abstract-We consider in this paper packets which arrive according to a Poisson process into a finite queue. A group of consecutive packets forms a frame (or a message) and one then considers not only the quality of service of a single packet but also that of the whole message. In order to improve required quality of service, either on the frame loss probabilities or on the delay, discarding mechanisms have to be used. We analyze in this paper the performance of the Early Message Discard (EMD) policy at the buffer, which consists of (1) rejecting an entire message if upon the arrival of the first packet of the message, the buffer occupancy exceeds a threshold ¤ , and (2) if a packet is lost, then all subsequent arrivals that belong to the same message are discarded.Index Terms-EMD policy, packet model, queue-length distribution, goodput. I. INTRODUCTIONQuite often quality of service have to be studied with respect to not only a single packet, but to a whole message or a frame. For example, in ATM a transport layer protocol (AAL) is responsible for grouping packets into a frame, and a lost packet implies the corruption of the whole frame. Selective Message Discarding (and EMD in particular, on which we focus here) have been proposed to achieve the twin goals of increased goodput and reduced network congestion by discarding the packets which do not belong to (or have potentials of not belonging to) good messages (a message is good if it is entirely received at the destination). Rejecting entire messages could also serve to guarantee an acceptable average delay bound for accepted messages. The goal of this paper is to present explicit expressions for the queue-length distribution and the goodput (defined as in [10] as the ratio between total packets comprising good messages exiting the network node and the total arriving packets at the input). Our starting point is the Markovian model proposed in [10]: a Poisson process of packet arrivals, geometrically distributed frame size, and exponentially distributed service times of packets. In [10], recursive procedures have been proposed for the computation of the performance measures, but explicit expressions have not been obtained. Our analytical results on closed form expressions for performance metrics (in particular the queue-length distribution and the goodput) may be quite useful in dimensioning the buffer size that should be used for a given goodput, in the study of the sensitivity of the goodput to different parameters for e.g., the message length, the buffer size, the load and most importantly in finding an estimate of the optimal discarding threshold etc.In a previous work [5], we analyzed the Partial Message Discard (PMD) policy in which only if some packet of a message is lost, subsequent packets are rejected (but entire messages are not discarded, in contrast with EMD). As the packet level analysis turns to be quite complex and involved, we studied in [3],

In this paper, we investigate the use of field programmable gate arrays (FPGAs) to accelerate relational joins. Relational join is one of the most CPU-intensive, yet commonly used, database operations. Hashing can be used to reduce the time complexity from quadratic (naïve) to linear time. However, doing so can introduce false positives to the results which must be resolved. We present a hash-join engine on FPGA that performs hashing, conflict resolution, and joining on a PCIe-attached system, achieving greater than 11x speedup over software.

No abstract