Patent Application
20140136668
The field relates to optimizations of software systems. More precisely, the field relates to real-time self-optimizing system for multitenant based cloud infrastructure.
One architecture pattern for large software systems consists of a persistency layer, an application layer, and a client layer. Typically, the persistence layer consists of a large database server, the application layer consists of many application servers, and the client layer handles at least thousands of users connected with client software such as, for example, an internet browser.
Today's challenge in the operation of such large software systems is to minimize the corresponding operational costs and to optimize the system to fulfill the expectations of the users regarding performance, reliability, security, and more. Most of the current approaches are based on the experience of the operators and the best practices which have been developed over a long period of time. The optimization and operation of those large software systems is a tedious trial.
Typically, large software systems consist of a persistency layer, an application layer, and a client layer which leads to different hardware requirements including different kind of configurations. The kinds of configuration have a huge variation from real physical hardware resources, operating system resource, up to more logical business application resources. Physical hardware resources are related to the used machine equipment like bus, memory, number and speed of central processor units (CPUs), network, hard disks, printers, etc. Operating system resources are related to virtual memory, caches, queues, semaphores, files, threads, processes, services, etc. Business application resources are related to business objects, agents, business tasks, mass data run objects, reports, etc.
For each machine responsible for the software execution on a specific layer an optimal list of all needed configuration parameters exists in a profile. In reality this profile is spread over many locations on a dedicated machine. Physical hardware parameters may be configured in the basic input/output system (BIOS) in case of a personal computer and the operating system resources may be configured in the Registry for example in case of the Microsoft Windows operating system. In addition, the business application resources may be configured in the business application system. Many of these parameters in the profile have dependencies between each other and cannot be configured without any constraints. For example, the maximum size of virtual memory depends on the available physical memory and the available disk space. In addition, the maximum table space size of a database also depends on the available disk space as well as the table caches depends on the available size of physical memory. Due to the large number of configuration parameters and their constraints, the optimization of a large software system is a challenge.
Various embodiments of systems and methods for real-time self-optimization of a business software system on multitenant based cloud infrastructure are described herein. In one embodiment, the method includes receiving an initial configuration of one or more application servers of a productive stream of the multitenant based cloud infrastructure, the productive stream operating on a tenant-enabled database, and wherein the initial configuration is within a freespace configuration defining ranges of configuration parameters based on key performance indicators (KPIs) for the performance of the business software system. The method further includes creating a simulation stream to operate parallel to the productive stream for configuration optimization purposes, the simulation stream comprising a tenant clone of the tenant-enabled database and a set of application servers reserved for simulations arid setting a simulation configuration by varying configuration parameters within the defined ranges of configuration parameters for the set of application servers reserved for simulation. The method further includes receiving one or more user requests to the tenant-enabled database through the productive stream, the one or more user requests processed by the one or more application servers of the productive stream and dispatching the received one or more user requests to the tenant clone of the tenant-enabled database through the simulation stream, the one or more user requests processed by the set of application servers reserved for simulations. The method further includes monitoring performance of the productive stream and the simulation stream for a predefined period and amount of processed user requests and applying the simulation configuration to the productive stream based on a comparison of KPIs of the simulation stream and KPIs of the productive stream.
In some embodiments, the system includes a tenant-enabled database and one or more application servers operating on the tenant-enabled database, the tenant enabled database and the one or more application servers forming a productive stream. The system also includes a tenant clone of the tenant-enabled database and a set of simulation application servers reserved for simulations, the simulation application servers operating on the tenant clone of the tenant-enabled database, the tenant clone and the set of simulation application servers forming a simulation stream. The system further includes a load balancer to handle client requests to the one or more application servers operating on the tenant-enabled database and to dispatch the client requests to the tenant clone of the tenant-enabled database through the set of simulation application servers. The system also includes an optimizer to set a simulation configuration by varying configuration parameters within defined ranges of configuration parameters for the set of simulation application servers, monitor performance of the tenant-enabled database with the one or more application servers and the tenant clone of the tenant-enabled database with the set of simulation application servers for predefined period and amount of processed client requests, and apply the simulation configuration to the productive stream based on a comparison of key performance indicators (KPIs) of the simulation stream and KPIs of the productive stream.
These and other benefits and features of embodiments will be apparent upon consideration of the following detailed description of preferred embodiments thereof, presented in connection with the following drawings.
The claims set forth the embodiments with particularity. The embodiments are illustrated by way of examples and not by way of limitation in the figures of the accompanying drawings in which like references indicate similar elements. The embodiments, together with its advantages, may be best understood from the following detailed description taken in conjunction with the accompanying drawings.
Embodiments of techniques for real-time self-optimizing system for multitenant based cloud infrastructure are described herein. In the following description, numerous specific details are set forth to provide a thorough understanding of the embodiments. One skilled in the relevant art will recognize, however, that the embodiments can be practiced without one or more of the specific details, or with other methods, components, materials, etc. In other instances, well known structures, materials, or operations are not shown or described in detail.
Reference throughout this specification to “one embodiment”, “this embodiment” and similar phrases, means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one of the one or more embodiments. Thus, the appearances of these phrases in various places throughout this specification are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
At block 220, a simulation stream is created to operate parallel to the productive stream for configuration optimization purposes. The simulation stream is dedicated to testing purposes simulating the real productive environment. In one embodiment, the simulation stream includes a tenant clone of the tenant-enabled database and a set of application servers reserved for simulation. The tenant clone of the tenant-enabled database is an exact copy of the tenant enabled database and real requests may be redirected to the tenant clone for testing purposes.
At block 230, a simulation configuration is set by varying configuration parameters within defined ranges for the set of application servers reserved for simulation. The ranges of configuration are defined by the freespace configuration, as depicted in connection to block 210.
Further, at block 240, user requests are received to the productive stream and the user requests are dispatched to the simulation stream. The received user requests to the productive stream are requests to tenant-enabled database and processed by the application servers of the productive stream. The dispatched user requests to the simulation stream are requests processed by the application servers reserved for simulation. The application servers reserved for simulation operate on the tenant-clone of the tenant enabled database and hence the user requests may affect the content both of the tenant-enabled database and the tenant clone at the same time. In one embodiment, a predefined percentage of the received user requests are dispatched to the tenant clone of the tenant-enabled database. In one embodiment, a percentage at the range of 5-10 from the total received requests is dispatched to the simulation stream, thus sealing the real requests for testing purposes.
Then, at block 250, the performance of the productive stream and the simulation stream are monitored for a predefined period and amount of processed user requests. In one embodiment, monitoring the performance of the productive stream and the simulation stream includes comparing database images of the tenant-enabled database and the tenant clone of the tenant-enabled database for inconsistencies. Inconsistencies of the compared images determine not equally processed user requests by the productive stream and the simulation stream. This means the simulation stream has not processed the requests the same way as the productive stream, which causes different data in the tenant-enabled database and the tenant clone of the tenant-enabled database. In another embodiment, the simulation stream is aborted, when the compared database images are inconsistent.
At block 260, a comparison is performed, whether the KPIs of the simulation stream are better than the KPIs of the productive stream. For example, if a KPI is “minimum response time” and the response time of the simulation stream is shorter than the response time of the productive stream, this determines that the KPI of the simulation stream is better the one of the current productive stream. Another example is if KPI is “minimum number of servers”, then the KPI of the simulation stream is better than the KPI of the productive stream if the simulation stream performs the same load of user requests with fewer servers than the simulation stream.
If the KPI's of the simulation stream are better than the KPIs of the productive stream, this means the simulation configuration set at block 230 is better than the current configuration of the productive stream and the process continues at block 270 with applying the simulation configuration to the productive stream and then turning back to block 230 for setting another simulation configuration for testing. If the KPI's of the simulation stream are not better than the KPIs of the productive stream, this means the simulation configuration set at block 230 is not better than the current configuration of the productive stream, the process gets back to block 230 setting another simulation configuration by varying configuration parameters within the defined ranges. Then another simulation starts by dispatching real user requests to the newly configured simulation stream running in parallel to the productive stream. In one embodiment, applying the simulation configuration to the productive stream when the KPIs of the simulation stream are better than the KPIs of the productive stream is performed by applying configuration parameters in steps and observing the performance of the business software system. By applying the configuration parameters in steps is aimed to adjust the productive system configuration gradually. Gradual implementation of configuration parameters allows tracking of configuration parameters that may cause undesired performance of the productive stream.
In one embodiment, a backup is created of the configuration of the productive stream. The back-up is used for rollback purposes in case an applied configuration to the productive stream results in the productive stream performing in an undesired way.
The system 300 also includes a load balancer 350 to handle client requests 355 from clients 360 to the one or more application servers 320 operating on the tenant-enabled database 310 and to dispatch the received one or more client requests 355 to the tenant clone 330 of the tenant-enabled database 310 through the set of simulation application servers 340. In one embodiment, the load balancer 350 is operable to dispatch a predefined percentage of the client requests 355 to the tenant clone 330 of the tenant-enabled database 310.
The system 300 further includes an optimizer 370. The optimizer 370 is operable to set a simulation configuration by varying configuration parameters within defined ranges of configuration parameters for the set of simulation application servers 340. The defined ranges of configuration parameters are persisted in a freespace configuration 377 within the optimizer 370. The defined ranges are based on key performance indicators (KPI 373) for the performance of the business software system 300. The optimizer 370 is further operable to monitor performance of the tenant-enabled database 310 with the one or more application servers 320 and the tenant clone 330 of the tenant-enabled database 310 with the set of simulation application servers 340 for predefined period and amount of processed client requests 355. For that purpose, the optimizer 370 processes simulation and baseline data 375. The optimizer 370 is further operable to apply the simulation configuration to the productive stream 315 when key performance indicators (KPIs) of the simulation stream 335 are better than KPIs of the productive stream 315. In one embodiment, the optimizer 370 is operable to compare database images of the tenant-enabled database 310 and the tenant clone 330 of the tenant-enabled database 310 for inconsistencies. In another embodiment, the optimizer 370 is operable to abort the simulation stream 335 when the compared database images are inconsistent.
In one embodiment, the optimizer 370 is operable to apply the configuration parameters in steps and observe performance of the business software system when applying the simulation configuration to the productive stream 315. In yet another embodiment, the optimizer 370 is operable to create a back-up of the configuration of the productive stream 315 for rollback purposes.
A real-time self-optimizing system for multitenant based cloud infrastructure such as the system 300 presented in
Another example of using a real-time self-optimizing system for multitenant based cloud infrastructure such as the system 300: there is a system with 10 application servers, 50 tenants, and 50,000 users. The goal is to minimize the size of the session roll areas (memory). The free-space configuration is set to 4 application server as simulation servers, the maximum number of simulation tenants is 5, the maximum number of simulation users is 1000, and the simulation shall run for 5 working days. The simulation tenants and simulation users will be determined by random. Absolute minimum and absolute maximum for session roll area size is configured. Tenant-copies will be created for each simulation pass. Each pass runs for one working day. The optimizer uses 5 passes to find the best fit session roll area size starting with min size and a step size of (max size−min size)/5. The simulation servers will be configured with this session roll area sizes and the requests of the simulation tenant users will be additionally routed to the simulation servers. The response time will be monitored. The optimized size of the session roll area is the result of the best fit number of the simulations.
Some embodiments may include the above-described methods being written as one or more software components. These components, and the functionality associated with each, may be used by client, server, distributed, or peer computer systems. These components may be written in a computer language corresponding to one or more programming languages such as, functional, declarative, procedural, object-oriented, lower level languages and the like. They may be linked to other components via various application programming interfaces and then compiled into one complete application for a server or a client. Alternatively, the components maybe implemented in server and client applications. Further, these components may be linked together via various distributed programming protocols. Sonic example embodiments may include remote procedure calls being used to implement one or more of these components across a distributed programming environment. For example, a logic level may reside on a first computer system that is remotely located from a second computer system containing an interface level (e. g., a graphical user interface). These first and second computer systems can be configured in a server-client, peer-to-peer, or some other configuration. The clients can vary in complexity from mobile and handheld devices, to thin clients and on to thick clients or even other servers.
The above-illustrated software components are tangibly stored on a computer readable storage medium as instructions. The term “computer readable storage medium” should be taken to include a single medium or multiple media that stores one or more sets of instructions. The term “computer readable storage medium” should be taken to include any physical article that is capable of undergoing a set of physical changes to physically store, encode, or otherwise carry a set of instructions for execution by a computer system which causes the computer system to perform any of the methods or process steps described, represented, or illustrated herein. Examples of computer readable storage media include, but are not limited to: magnetic media, such as hard disks, floppy disks, and magnetic tape; optical media such as CD-ROMs, DVDs and holographic devices; magneto-optical media; and hardware devices that are specially configured to store and execute, such as application-specific integrated circuits (“ASICs”), programmable logic devices (“PLDs”) and ROM and RAM devices. Examples of computer readable instructions include machine code, such as produced by a compiler, and files containing higher-level code that are executed by a computer using an interpreter. For example, an embodiment may be implemented using Java®, C++, or other object-oriented programming language and development tools. Another embodiment may be implemented in hard-wired circuitry in place of, or in combination with machine readable software instructions.
A data source is an information resource. Data sources include sources of data that enable data storage and retrieval. Data sources may include databases, such as, relational, transactional, hierarchical, multi-dimensional (e. g., OLAP), object oriented databases, and the like. Further data sources include tabular data (e. g., spreadsheets, delimited text tiles), data tagged with a markup language (e. g., XML data), transactional data, unstructured data (e. g., text files, screen scrapings), hierarchical data (e. g., data in a file system, XML data), files, a plurality of reports, and any other data source accessible through an established protocol, such as, Open DataBase Connectivity (ODBC), produced by an underlying software system (e. g., ERP system), and the like. Data sources may also include a data source where the data is not tangibly stored or otherwise ephemeral such as data streams, broadcast data, and the like. These data sources can include associated data foundations, semantic layers, management systems, security systems and so on.
In the above description, numerous specific details are set forth to provide a thorough understanding of embodiments. One skilled in the relevant art will recognize, however that the embodiments can be practiced without one or more of the specific details or with other methods, components, techniques, etc. in other instances, well-known operations or structures are not shown or described in details.
Although the processes illustrated and described herein include series of steps, it will be appreciated that the different embodiments are not limited by the illustrated ordering of steps, as some steps may occur in different orders, some concurrently with other steps apart from that shown and described herein. In addition, not all illustrated steps may be required to implement a methodology in accordance with the one or more embodiments. Moreover, it will be appreciated that the processes may be implemented in association with the apparatus and systems illustrated and described herein as well as in association with other systems not illustrated.
The above descriptions and illustrations of embodiments, including what is described in the Abstract, is not intended to be exhaustive or to limit the one or more embodiments to the precise forms disclosed. While specific embodiments of, and examples for, the present disclosure are described herein for illustrative purposes, various equivalent modifications are possible within the scope of the present disclosure, as those skilled in the relevant art will recognize. These modifications can be made in light of the above detailed description. Rather, the scope is to be determined by the following claims, which are to be interpreted in accordance with established doctrines of claim construction.
