Croc: System of complex monitoring

The Croc company provided the system of complex monitoring available as in a format of the managed service from a cloud Croc, and for deployment on own platform.

Appointment

Several monitoring systems which or are delivered together with hardware complexes quite often are responsible for efficiency of IT in the large company, or are implemented pointwise for specific sites of infrastructure. Information from these systems is separate and is stored only in their own databases. As a result of it it is impossible to use for complex analytics which will help to define the root reasons of problems in IT.

The solution Croc for umbrella monitoring allows to use as much as possible already available components of monitoring and their data for ensuring smooth operation of IT services.

The System of Complex Monitoring (SCM) is created as the answer to 3 calls which domestic customers – foreign sanctions, technological limits and high cost of ownership face.

Foreign sanctions force customers to consider in all IT fields the alternative or domestic solutions and services allowing them to receive a certain political protection – as in the present, and in the future. In addition, many customers, having wide experience of implementation of large software packages, begin to understand that big "box", the ready-made solution from industrial vendor, not always allows to receive necessary functions, flexible architecture, without speaking about volume, that not all equipment or software is supported. They should bypass somehow independently different technological limits or to work only within available opportunities. The third call – the price. After 2014 the cost of ownership of foreign software considerably raised and increases still.

These three aspects became the reasons and import substitution programs, and search of developments of open source to build a flexible, effective and vendoronezavisimy solution. In practice Croc they poured out in a row RnD-projects and emergence of solutions on the basis of which industrial solutions can receive the software packages equivalent on functionality. Knowledge gained for the twenty-five year history of interaction with large vendors, their software products – architecture, approaches, technologies – allowed Croc company to create similar complexes, but already on open components.

The system of complex monitoring integrates more than 10 products with the open code, providing the maximum flexibility of integration with elements to an infrastrukt

]]

Criminal Police Service – one of such examples. Understanding IT architecture, knowing the current perspective of the customer and having broad examination in many areas, Croc developed own solution stack and technologies for complex monitoring of IT.

Functionality

1. Infrastructure monitoring – monitoring of servers, network equipment, disk subsystems, the engineering equipment, infrastructure and web services, etc., tracking of operability of applications and elements of applications.

2. Transaction monitoring – what volume of useful traffic takes place on network? What transactions contain in useful traffic? How they are expedited? Whether the user transactions in the interface quickly work?

3. Monitoring of business services – of what configuration units consists IT infrastructure? What of them influence work of business services? How did high loading of the processor, failure of a disk in the disk array or unavailability of the router affect indicators of the company?

Levels of collecting, processing and data visualization

Criminal Police Service has the modular architecture allowing to select necessary modules under specific requirements of the customer at three logic levels. The first – the level of collection of data on functioning of IT infrastructure. Can be sources of these data as the installed tools, and systems which are already available for the customer: accounting, management systems, even some self-written solutions.

At the same time we are interested in two global data types: events about failures in these or those systems or objects of monitoring and functioning indicators – for example, parameters of availability, performance, etc. This level is modified depending on an IT landscape with which we work, and those systems which the customer has, and it can be changed most flexibly, - Valentin Nyk, the head of department of management systems for IT infrastructure explains Croc.

The following logic level, the most difficult, but also changed – processing and aggregation of data. The data array which arrived from external systems is intellectually processed and takes the form of ready analytical information: for example, it allows to find the failure prime cause, anomaly of functioning of IT infrastructure. At this level different practices Croc join: analytical engines, elements of machine learning, creation of resource and service models on the basis of which difficult correlation rules can be carried out difficult analytics.

Creation of adequate resource and service model – very uncommon problem, logic of calculation in it rather difficult. Earlier this function was really performed only by the big closed vendor solutions. But now Croc is able to make similar calculations too, and on the open source software.

Besides, we use machine learning in the field of monitoring systems. It is no secret that for adequate creation of behavior model of an object large volumes, and monitoring — a good source are necessary for this, in fact, Big Data. We collect information on functioning of infrastructure facilities to do difficult analytics which allows to achieve, for example, forecasting of indicators of IT infrastructure or service, - Valentin Nyk adds.

Monitoring data processing by the engine of machine learning allows model to study identification of causes and effect relationships independently. It will prediktivno allow or to react to a problem, or reaktivno to solve it in case of emergence. The saved-up knowledge base about possible solutions of a problem helps the operator of a system to select the most suitable. Practice shows that such approach allows to reduce considerably recovery time after failure. It is, in fact, a step to the future — when engines of machine learning will predict problems prediktivno or reaktivno to solve them in the automatic mode using instruments of automation, being based based on knowledge. The future of monitoring systems behind intellectual event handling.

The third level – visualization of already processed information: different dashborda, reports – a peculiar situational center. Here Croc offers the intuitive, functional design individual for each customer, using open frameworks with a ready set of different graphic widgets and portlets.

Advantages over proprietary products

Independence of vendor in emergency situations. Waiting of the new version of a proprietary product can last for months, custom completions result in unpredictable behavior of a rabotyvsy software package and growth of the IT budget.

Modular architecture. The solution Croc is developed taking into account the existing market situation and allows to be integrated with any products of the leading developers. At the same time the customer can pick up those functions which are relevant for him and to refuse unnecessary.

High customization of the interface. When using the proprietary systems the customer adapts to the preconfigured indicators, often without receiving display detailing necessary to it.

Simple integration in the existing IT landscape. Proprietary products can not have tools necessary for the customer for data collection. It threatens all monitoring as such and leads to additional expenses on completion of the solution. At the same time a part of necessary functionality can remain unavailable.

In case of use of the solution Croc at data collection we are not limited in sources or the maintained technologies — as are ready that many things will need to be finished under the customer. We are initially open for expansion of functionality and inclusion in a system of certain, originally not supported data sources, - Valentin Nyk explains.

Use options

Customers can use the system of intellectual monitoring from a cloud Croc in a format of the managed service or unroll it on own platform. Service allows to transfer capital costs to OPEX and also to manage flexibly functionality of a system and to select only those opportunities which are really necessary for the company. Due to use the customer receives solutions from a cloud a ready-made product in 2 weeks after agreement signature.

Ownership cost

Researches Croc showed that the cost of ownership (TCO) of a foreign product where the cost of licenses enters, supports, services in its deployment and setup on average more than twice above, than the implementation cost of a stack of technologies which developed Croc.

Actually, implementation of vendor solutions of monitoring are too expensive to many companies SMB, and now situations only worsens. They or looked for products of some niche players with the reasonable price label, but small functionality, or used the separate open source components, increasing own competence. We come with a complete solution stack which will allow to close all business needs, but at the same time to consider possibilities of each customer individually. Each logic level of Criminal Police Service can be functionally simplified or, on the contrary, is expanded — and its cost will depend on it. Therefore the solution is suitable including for a segment of SMB, under projects within several million rubles, - Valentin Nyk, the head of department of management systems for IT infrastructure tells Croc.