Maheswaran Surendra

Controlling Quality of Service in Multi-Tier Web Applications

The need for service differentiation in Internet services has motivated interest in controlling multi-tier web applications. This paper describes a tier-to-tier (T2T) management architecture that supports decentralized actuator management in multi-tier systems, and a testbed implementation of this architecture using commercial software products. Based on testbed experiments and analytic models, we gain insight into the value of coordinated exploitation of actuators on multiple tiers, especially considerations for control efficiency and control granularity. For control efficiency, we show that more effective utilization of tiers can be achieved by using actuators on the bottleneck tier rather than only using actuators on the entry tier. For granularity of control (the ability to achieve a wide range of service level objectives) we show that a fine granularity of control can be achieved through a coordinated, cross-tier exploitation of coarse grained actuators (e.g., multiprogramming level), an approach that can greatly reduce controllerinduced variability.

Radiofrequency discharge benchmark model comparison

A well defined basis of comparison for self-consistent simulations of RF discharges is presented. Twelve different simulation efforts including kinetic, hybrid and fluid methods have produced results for a one-dimensional RF discharge simulation using identical cross sections or swarm parameters under specified conditions. In addition, global model predictions are also provided for comparison. The agreement between groups using the particle-in-cell with Monte Carte collisions (PIC-MCC) technique and a hybrid method is good. Differences between PIC-MCC and convective scheme solutions to the Boltzmann equation exist primarily in the electron number density and not energy density. Fluid simulations generally show greater variation in the results than kinetic/hybrid schemes. The differences, which are typically greatest at low pressures, are apparent only in some plasma parameters, such as RF voltage and plasma density, but not in others, such as ion flux to the electrodes. In general, the agreement between global model results and simulation predictions is quite reasonable. These results provide some indication of the sensitivity of the solutions to approximations and numerical techniques used in the various models.

Dynamic Surge Protection: An Approach to Handling Unexpected Workload Surges with Resource Actions that Have Lead Times

Today’s information technology departments have significantly varying demands for resources due to unexpected surges in subscriber demands (e.g., a large response to a product promotion). Further complicating matters is that many resource actions done in response to surges (e.g., provisioning or de-provisioning an application server) have substantial delays (lead times) between initiating the resource action and its taking effect. This paper describes dynamic surge protection, an approach to handling unexpected workload surges in systems that have lead times for resource actions. Dynamic surge protection incorporates three technologies: adaptive short-term forecasting, on-line capacity planning, and configuration management. The paper includes empirical results from evaluations done on a research testbed, including favorable comparisons with a threshold-based heuristic. The results from an extended test also show that service objectives can be maintained cost-effectively.

Incorporating cost of control into the design of a load balancing controller

Load balancing is widely used in computing systems as a way to optimize performance by reducing bottleneck utilizations, such as adjusting the size of buffer pools to balance resource demands in a database management system. Load balancing is generally approached as a constrained optimization problem in which only the benefits of load balancing are considered. However, the costs of control are important as well. Herein, we study the value of including in controller design the trade-off between the cost of transient imbalances in resource utilizations and the cost of changing resource allocations. An example of the latter are actions such as resizing buffer pools that can reduce throughputs. This is because requests for data in pools whose memory is reduced immediately have longer access times whereas requests for data in pools whose memory is increased must fill this memory with data from disk before accessed times are reduced. We frame our study of control costs in terms of the widely used linear quadratic regulator (LQR). We develop a cost model that allows us to specify the LQR Q and R matrices based on the impact on system performance of changing resource allocations and transient load imbalances. Our studies of a DB2 universal database server using benchmarks for online transaction processing and decision support workloads show that incorporating our cost model into the MIMO LQR controller results in a 14% improvement in performance beyond that achieved by dynamically allocating the size of buffers without properly considering the cost of control.

Generic on-line discovery of quantitative models for service level management

Quantitative models are needed for a variety of management tasks, including (a) identification of critical variables to use for health monitoring, (b) anticipating service level violations by using predictive models, and (c) on-going optimization of configurations. Unfortunately, constructing quantitative models requires specialized skills that are in short supply. Even worse, rapid changes in provider configurations and the evolution of business demands mean that quantitative models must be updated on an on-going basis. This paper describes an architecture and algorithms for on-line discovery of quantitative models without prior knowledge of the managed elements. The architecture makes use of an element schema that describes managed elements using the common information model (CIM). Algorithms are presented for selecting a subset of the element metrics to use as explanatory variables in a quantitative model and for constructing the quantitative model itself. We further describe a prototype system based on this architecture that incorporates these algorithms. We apply the prototype to on-line estimation of response times for DB2 Universal Database under a TPC-W workload. Of the approximately 500 metrics available from the DB2 performance monitor, our system chooses 3 to construct a model that explains 72% of the variability of response time.

Generic Online Optimization of Multiple Configuration Parameters with Application to a Database Server

Optimizing configuration parameters is time-consuming and skills-intensive. This paper proposes a generic approach to automating this task. By generic, we mean that the approach is relatively independent of the target system for which the optimization is done. Our approach uses online adjustment of configuration parameters to discover the system’s performance characteristics. Doing so creates two challenges: (1) handling interdependencies between configuration parameters and (2) minimizing the deleterious effects on production workload while the optimization is underway. Our approach addresses (1) by including in the architecture a rule-based component that handles interdependencies between configuration parameters. For (2), we use a feedback mechanism for online optimization that searches the parameter space in a way that generally avoids poor performance at intermediate steps. Our studies of a DB2 Universal Database Server under an e-commerce workload indicate that our approach can be effective in practice.

A configuration management database architecture in support of IBM service management

In this paper, we present the architecture of the IBM Tivoli® Change and Configuration Management Database. Its main features include a rich data model, automatic discovery of data for configuration items, visualization of application dependencies on configuration items, and multicustomer support. We discuss implementation topics, such as relationship management, composite configuration items, data federation, reconciliation of data from different sources, a security model for multicustomer support, and integration of change-management and configuration-management processes.

Self-consistent fluid modeling of radio frequency discharges in two dimensions

Results from a two‐dimensional (2D) fluid simulation of a parallel plate, capacitively coupled radio frequency discharge bounded by a cylindrical insulator with a grounded exterior surface are presented. We find that the radial sheath at the insulator focuses current into the plasma region adjacent to the sheath. This 2D effect has important ramifications for the ionization rate, which peaks sharply in the metal‐insulator corners. We have experimentally observed the enhancement of the emission rate in a corner using spatially resolved optical emission spectroscopy. A ‘‘thick’’ insulator yields radial profiles for the time‐averaged plasma density and potential that are essentially uniform. A ‘‘thin’’ insulator, however, results in an off‐axis maximum in the plasma density and potential due to the corner ionization.

Modeling Differentiated Services of Multi-Tier Web Applications

In this paper we present a hybrid performance model for modeling differentiated service of multi-tier web applications with per-tier concurrency limits, cross-tier interactions, as well as a work-conserving resource allocation model. The service dependencies between multiple tiers are captured first using a layered queueing model. We then show how to model per-tier concurrency limits and service differentiation between multiple classes while maintaining work conservation at each tier. We use a function approximation approach combined with a coupled processor model. Our model is calibrated from an actual multitier J2EE testbed, and we show the ability of the model to accurately model common performance metrics. Our proposed (layered) model shows 78% improvement in root mean square error over a single-tier machine repair model as well as a tandem queue model. We also demonstrate one application of the model for model-based resource allocation.

Control Theory: a Foundational Technique for Self Managing Databases

Control theory is a well established discipline that has emerged from aeronautical, electrical, and mechanical engineering to provide a formal approach to building robust systems. While similar robustness concerns exist in database management systems, control theory is rarely used due to the lack of canonical control models and a dearth of control theory expertise among database researchers. We discuss our experience with using control theory to build self managing databases, showing experimental results, discussing pitfalls and limitations, and contrasting formal models against with feedback loops. While our experience indicates that control theory is a good paradigm for database self management, control theory should be used Judiciously since its techniques are not suited to all problems in database administration.