=Paper=
{{Paper
|id=Vol-1787/167-171-paper-28
|storemode=property
|title=Automation of organizational and technical arrangements for scientific researches
|pdfUrl=https://ceur-ws.org/Vol-1787/167-171-paper-28.pdf
|volume=Vol-1787
|authors=Alexander Degtyarev,Andrei Ivashchenko,Vladimir Korkhov,Nikolai Yuzhanin
}}
==Automation of organizational and technical arrangements for scientific researches==
https://ceur-ws.org/Vol-1787/167-171-paper-28.pdf
Automation of organizational and technical
arrangements for scientific researches
A.B. Degtyareva, A.I. Ivashchenkob,
V.V. Korkhovc, N.V. Yuzhanind
Saint Petersburg State University, 7/9 Universitetskaya emb., St. Petersburg, 199034, Russia
E-mail: adeg@csa.ru, baiivashchenko@cc.spbu.ru, cvladimir@csa.ru, dyuzhanin@cc.spbu.ru.
The article highlights the set of problems associated with automation of computing and informational sup-
port of scientific research, and offers a possible integrated solution based on the service desk system. The atten-
tion is focused on the common organizational and technical activities related to the lifecycle of scientific re-
search, such as registration, accounting and technical support. A gracefully integrated software complex with the
usage of universal web service is provided as a solution. Mentioned system binding tool allows automating the
workflow of key applications and simplifying the staff decision-making problem. Achieved data relevance, re-
duction of human impact factor and man-hour costs are mentioned as a positive factor of integration solution.
Keywords: ITSM, Service Desk, Integration, Infrastructure, Computing Center.
Research was carried out using computational resources provided by Resource Center "Computer Center of
SPbU"(http://cc.spbu.ru/) and supported by grants of Russian Foundation for Basic Research (projects no. 16-07-01111,
16-07-00886, 16-07-01113) and Saint Petersburg State University (project no. 0.37.155.2014).
© 2016 Alexander B. Degtyarev, Andrei I. Ivashchenko, Vladimir V. Korkhov, Nikolai V. Yuzhanin
167
Introduction
Today’s university computing center is a complex infrastructure with multiple interdependencies
between its components no matter how large its scale is. Support of such infrastructure is absolutely
non-trivial activity. A university specific in case of scientific researches is in variety of scientific prob-
lems that is a consequence of the variety of faculties, departments and laboratories. To provide IT ser-
vices to all those users with different requirements the computing center is forced to support different
resources and methodic of service. These requirements cause three basic problems opposed to high
quality of service and user satisfaction: large amount of human tasks in processes of support, difficul-
ties with construction of unified accounting system, and dependency of the data relevance from human
impact factor. Solution of these contradictions is also complicated by strong limitations on financial
and human resources that are inherent for governmental universities. In our case we tried to solve it
with “low blood” solution represented with combination of informational and organizational solutions.
Main approach
As an approach to the construction of our solution we identified four basic activities in scientific
research support mostly exposed to the problems mentioned before. Next step was in consideration of
these activities in the scope of the ITIL processes these activities are related to (Pic.1). The ITIL scope
was chosen because of its compliance to the standards of quality of IT services, and also because most
of software solutions for IT service management is certified on compliance to this library of best prac-
tices.
Pic.1. Activities and processes of support
On the theoretical step we concluded that full automation of all these processes is unattainable
because of university specific: center needs to support a lot of too different researches. That’s why we
found it necessary to combine information system and organizational solution. To simplify these or-
ganizational arrangements we decided to try auxiliary information system. The goal of the auxiliary
system is to decrease the impact of human tasks on service quality. In our case it is expressed in deci-
sion support for the computing center staff.
On the other hand to keep all that subsystems united we must use some system with common ob-
jects and interface. In such case the service desk system seems like a best choice because of several
reasons. First reason is a framework that provides a flow of documents for all necessary processes of
support. But this flow mostly manual and this fact cannot satisfy the computing center staff. Second
reason is in the fact that service desk has its own data store with basic object necessary for support
168
processes: tickets, FAQ notes, etc. Third reason is a web service API included in all widely known
service desk frameworks.
Because of many additional systems we are planning to connect with service desk it must be cus-
tomizable. That is why it must be open-source solution. And there is only one open-source service
desk that meets all the requirements – the OTRS::ITSM service desk system. Its key features are fully
customizable framework, included CMDB, web service API and automated report generator.
Next key element is usage of web services. It provides necessary integration of service desk and
additional systems, and also the external information systems no matter higher or lower level they are.
Web services are easy to use. But they have some disadvantages: different message formats and varie-
ty of data structures.
Problems and solutions
Main problem we met was unification of the system-to-system communication. To solve this
problem the set of web service is used. Solution of this problem led us to develop some sort of consol-
idation API (Pic.2) to organize and automate the data exchange between information systems.
Pic.2. Structure of support system
Second significant problem was data control for error prevention. Some manipulations with the
data in processes of research supports remained manual so the processed data must be checked before
saving. This feature was included in our automation system we called Cougar. And this feature be-
came one of factors of human impact factor limitation.
Automation complex
One of the challenges we met during dataflow has been built is an integration of various informa-
tional systems and services. This problem appears due to the fact that there is no complex and unified
solution on the market, and we are even not talking about open source. The most flexible systems pro-
vide a basic functionality out of the box, like ticket management, workflow organization, automatic
169
call registration etc., but the case management for the common lifetime processes is an end user's re-
sponsibility. Commercial development of such case-based solutions is very expensive and due to that
fact usually not available for university structures.
Keeping all that in a mind we have developed our own service, called "Cougar", which goal is to
orchestrate our informational and infrastructural systems in the scope of HPC-based scientific project
life cycle.
We realized it in the set of wizards for main processes of scientific research support. First step
was the description of main support procedures and turning those descriptions in sets of rules. The
wizards based on chains of those rules collected in a knowledge base. The system has a web-based UI
with a support of mobile devices. It is very friendly for the operators and allows them to work with a
system from any place where they have access to the web.
Pic.3. Cougar system interaction
Let's take a closer look for one of our common service scenarios, where the end user gets an ac-
cess to cluster workload management system through virtual machine. This process includes the fol-
lowing steps: creation of user account, granting required permissions to the user, creation of virtual
machine, sending a response to the ticket through service desk, logging everything to the CMDB. As
you can see on Pic. 3, there are a lot of services involved, and staff should know how to interact with
the whole stack, which also make a process of new employee acquirance quite painful and time con-
suming. Also, according to our stats, operators are often making mistakes, mostly with wrong input to
the CMDB. Right now they are not happening just because ticket is parsed automatically and operator
just need to check user input.
Conclusions
As a conclusion we can say that research led us to the creation of support system with automatic
advisory that significantly simplified the decision making for the computing center operators during
170
the processes of research lifecycle support. Thanks to knowledge base and wizards for main processes
this system reduced human impact factor and quantity of errors during operations with documents.
Relevance of the data in data stores also got its profit because automatic system cannot “forget” to
save, update, or delete necessary data.
Simplified report generation system became important acquirement for managers that reduced
complexity and time cost of periodic report creation. Also part of automated reports became available
for users and gave them an opportunity to evaluate their costs and efficiency during their research.
Perspectives
One of support problems was not solved completely – transport between the service desk system
and the monitoring and accounting systems still realized without usage of web service API. Solution
of this problem will be significant perspective work.
For the increase of service quality we want to automate the analysis of the users objectives, re-
quests and QA surveys data and provision of recommendations for users about usage of our services.
Such automation can reach two goals: to help in searching of optimal methodic to use in recent re-
search and either to suggest the proper resources to achieve the objectives of future researches.
References
Black J., Draper C., Lococo T., Matar F., Ward C. An integration model for organizing IT service
management // IBM Systems Journal. — 2007. — Vol. 46, No. 3. — P. 405–422.
Richardson L., Ruby S. RESTful Web Services. s.l. // O'Reilly. — 2007. — 454p.
Booth D., Haas H., McCabe F. Web Services Architecture. [Document] s.l. // W3C Working Group.
— 2004.
Bogdanov A., Gaiduchok V., Gankevich I., Tipikin Yu., Yuzhanin N. The use of service desk to keep
track of computational tasks on supercomputers // Transactions on Computational Science
XXVII. — 2016. — P. 1–9.
Ezhakova T., Gaiduchok V., Yuzhanin N., Zolotarev V. Development of the Configuration Manage-
ment system of the computer center and evaluation of its impact on the provision of IT services
// Computational Science and Its Applications – ICCSA 2015. — 2015. — Part. IV. — P. 249–
258.
Steinberg, R. ITIL Service Operation 2011 Edition (Best Management Practices) // London: The Sta-
tionery Office. — 2011. — 384p.
171