place via desired requirements. After choosing a
service he picks the best match and one of the ready-
made contracts that are offered. The customer sees
the costs for the service and can end and rate it.
Compared with the other approaches our conceptual
CloudMS supports the requirements of the first use
case best (Moltkau, 2012). While marketplace solu-
tions offer limited search mechanisms they rarely
support the monitoring of the service execution. The
provision of different agreements is not supported by
any of the existing systems.
In the second use case our conceptual system can
fulfill all requirements. The agreed service levels are
monitored and can be inspected at. When the re-
sponse time drops the customer is informed about
the attempt of the provider to adjust the service in-
frastructure. He can end the service and rate it. Dur-
ing the search for a new service he can increase the
priority for the response time. For the contract nego-
tiation the customer can use a form to set his desired
service quality. In comparison enStratus provides
“sufficient” and therefore the best support of the
existing systems. It does not support the matchmak-
ing of a service or the rating of the service perfor-
mance. None of the introduced systems support the
negotiation of an SLA or the setting of priorities
during the search.
The provider in the third use case can register a
service by submitting a service description. When a
consumer asks for executing the service the provider
starts the negotiation of an SLA. During service
execution the provider can adjust the infrastructure.
The costs and equalization payments are calculated
as well. The payment itself is handled by the system.
The provider can see the ratings of his service. All
requirements of the third case can be fulfilled by the
conceptual system. The AWS marketplace provides
“sufficient” and therefore the best support within the
existing systems but cannot forward alerts, monitor
the execution sufficiently, or adjust the service infra-
structure. The AWS marketplace provides “suffi-
cient” and therefore the best support within the ex-
isting systems but cannot forward alerts, monitor the
execution sufficiently, or adjust the service infra-
structure.
The conclusion of the evaluation is that the re-
quirements of the three use cases could be fulfilled
“excellent” by our conceptual CloudMS. The man-
agement of the last two use cases is possible in its
entirety and the first use case can be managed by the
system to nearly full extent. The existing CloudMS
support significantly less portions of the entire
lifecycle. The best result here is the support of the
third use case by the AWS Marketplace. Because
each of the systems lays its focus on a different area
they lack functionality when it comes to full lifecy-
cle support. The evaluation conveys that our concep-
tual CloudMD can not only satisfy the requirements
of one specific scenario but a broad field of man-
agement requirements.
6 FUTURE WORK
The drafted CloudMS is in a very early stage. The
next step is its evaluation by the target audience to
examine the usability of the system. The implemen-
tation of a prototype will follow to analyze the theo-
retical considerations in a practical employment. We
will examine if the recommendations are suitable for
daily use and if all requirements are convertible. The
shortcoming of the system regarding the level of
detail for the search should also be improved in the
future. Considerations on how the missing level of
detail can be achieved have to be included into the
system concept. Another improvement can be a
closer connection between the system components.
One example is the linking between the Service
Marketplace and the Controlling and Monitoring
Area. During the execution of a service the costs can
be monitored to offer cheaper services with similar
features.
REFERENCES
Amazon Web Services, LLC, 2012. AWS Management
Console. (2012a), AWS Marketplace. (2012b).
Baun, C., Kunze M., 2011. The KOALA cloud manage-
ment service. New York, USA.
Bizmanualz, Inc., 2011. How Are PDCA Cycles Used
Inside ISO 9001?
Bleizeffer, T., Calcaterra, J., Nair, D., Rendahl, R.,
Schmidt-Wesche, B., Sohn, P., 2011. Description and
Application of Core Cloud User Roles., USA.
Brandic, I., 2009. Towards Self-Manageable Cloud
Services.
Braun, I., Reichert, S., Spillner, J., Strunk, A., Schill, A.,
2008. Zusicherung nichtfunktionaler Eigenschaften
und Dienstgüte im Future Internet of Services. PIK.
enStratus Networks, Inc., 2012. enStratus Cloud
Management Overview.
Google, Inc., 2010. Google Apps Marketplace.
ISO/IEC Standard 27001, International Organization for
Standardization, 2005.
Joshi, K., Finin, T., Yesha, Y., 2009. Integrated Lifecycle
of IT Services in a Cloud Environment. NC.
Janiesch, C., Niemann, M., Steinmetz, R., 2011. The
TEXO Governance Framework, SAP Research.
CLOSER2013-3rdInternationalConferenceonCloudComputingandServicesScience
218