دليل تجارب الاسفلت المختبرية - Asphalt Experiments Guide Laboratory
Iaetsd pinpointing performance deviations of subsystems in distributed
1. Pinpointing performance Deviations of Subsystems in Distributed
Systems with cloud infrastructures
1
Hareesh Kamisetty M.Tech,
(Audisankara College Of Engineering And Technology,Gudur)
Abstract- Cloud Manufacturing Systems is
service oriented system (SOS) and they are used
to compose the various kinds of cloud related
applications. In which some of the applications
are failed to complete the execution of user
request with in the deadline. And the Cloud
Manufacturing Systems are still facing the
performance problems and the existing
techniques considering the tracing request data
as performance data to find and diagnosis the
performance of the service. In this paper, we
propose the Cloud Debugger. It is a promising
tool to diagnosis the performance problems in
Cloud Manufacturing Systems.
Index terms- Cloud Manufacturing System
(CMfg), Performance Diagnosis, Service-
Oriented.
I. INTRODUCTION
The Cloud Manufacturing Systems
service-oriented systems and compose the
various services. In the real world combining the
increased advanced technologies such as
virtualization, advanced computing technologies
and service-oriented technologies with existing
models and new wide range of information
technologies, a new computing and service
manufacturing model called cloud
manufacturing is proposed.
Four types of CMfg service platform
i.e., public private, hybrid and community CMfg
service platforms. Compared with existing
manufacturing models CMfg has the following
features or properties:
i) Service and requirement-oriented.
Most of the manufacturing models are
resource-or order-oriented, while CMfg is a
service-and requirement-oriented manufacturing
model. The core idea of CMfg is manufacturing
as a service.
ii) Dynamic with uncertainty.
The resources and services in CMfg are
various and dynamic, and the solutions for
addressing manufacturing tasks are dynamic.
iii) Knowledge-based.
The whole life cycle of CMfg system
needs knowledge support.
iv) Initiative.
In a Cloud manufacturing System, both
manufacturing requirements are active and they
can automatically find and match with each
other with the support of semantic reasoning
based on knowledge.
v) Physically distributed and logically
centralized
The physical developed resource and
facility in CMfg locate in different places and
323
INTERNATIONAL CONFERENCE ON CURRENT INNOVATIONS IN ENGINEERING AND TECHNOLOGY
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
ISBN: 378 - 26 - 138420 - 5
www.iaetsd.in
2. are restricted by different persons or
organizations.
a) Existing System
The art of debugging for cloud
applications is not much more than writing out
diagnostic messages and spelunking the logs for
them.
When the right data is not being write to
the logs, developers have to change the code and
redeploy the application to production. The
traditional debuggers are not well suited for
cloud-based services for two reasons. First, it is
hard to know which method to attach to. Second,
stopping a method in production makes it hard
to repeat an issue and gives end-users a bad
experience.
b) Proposed System
To address above mentioned problems
this paper proposes the a novel Debugger is
called as Cloud Debugger which significantly
reduce the effort of cloud operator to identify the
which method to attach and it provides a easy
way to repeat the service based on users
experiences.
II. PROPOSED WORK
The Cloud Debugger completely
changes this. It allows developers to start where
they know best-in the code. By simply setting a
watch point on a line of code, the next time a
request on any of the cloud servers hits that line
code, and we get a snapshot of all the local
variables, parameter, instance variables and a
full stack trace. There is zero setup time and no
complex configuration to enable. The debugger
is supreme model for utilize in production.
There is no overhead for enabling the debugger
on this paper.
a) Cloud Trace
Performance is an important feature of
service which directly interrelated with end
users satisfaction and maintenance. No on
intends to build a slow service, but it can be
extremely difficult to identify the core reasons of
slowness when it happens.
The cloud Trace helps you visualize and
understand the time spent by application for
request processing. This enables the cloud
operator to hastily find and repair performance
bottlenecks.
The cloud operator can easily produce a
report that shows the performance change in the
service from one release to another.
b) Cloud Monitoring
Cloud Monitoring provides the
dashboards and alerting capabilities that help the
cloud operator to identify and repair the
performance problem quickly.
With minimum configuration and no
separate infrastructure to maintain, Cloud
Monitoring provides deep visibility into Cloud
platform services when compared to the
traditional approaches.
Finally, we can create end point checks
to monitor availability and response times for
end-user facing services. Endpoint checks are
performed by problems in Texas, Virginia
enabling monitoring of latency.
324
INTERNATIONAL CONFERENCE ON CURRENT INNOVATIONS IN ENGINEERING AND TECHNOLOGY
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
ISBN: 378 - 26 - 138420 - 5
www.iaetsd.in
3. SSH to VM instantly
Sometimes it is unavoidable to connect
directly to a VM to debug or repair the
production issue. It can be a bit of pain,
especially when the operator is on the road, so
now we can do that from just about anywhere.
With our new browser based SSH client can
speedily and securely connect to any of VMs
from the console. It no needs to install any SDK
or any tools. This could be taken as future work.
III. CONCLUSION
The traditional approaches diagnosis the
performance problems based on the users
requesting trace data but it suffers from the
tradeoffs between tracing granularity and
debugging effort. As a result, more effort will be
increased in troubleshooting the performance
problems. So that, this paper proposes the Cloud
Debugger as tool for performance diagnosis with
effortless. And this can be significantly reducing
the communication and computation overheads.
FUTURE ENHANCEMENT
The proposed system of this paper
considers the cloud monitoring and cloud trace
as a techniques to find and repair the
performance problems. But the performance of
the service is also depends on the system
dynamics which is leave it as future research
work.
REFERENCES
[1] http://googlecloudplatform.blogspot.in / 20 -
14 /06 /enabling-developers-to-tame-product
-ion-systems-in-the-cloud.html.
[2] F, Tao; L Zhang; VC Venkatesh; YL Luo; Y
Cheng (2011). "Cloud manufacturing: a
computing and service-oriented manufactu -
ring model". Proceedings of the Institution of
Mechanical Engineers, Part B, Journal of
Engineering Manufacture. doi:10.1177/
0954405411405575
[3] Michael Rung-Tsong Lyu and Hua Cai
“Toward Fine-Grained, Unsupervised,
Scalable Performance Diagnosis for
Production Cloud Computing Systems” Proc.
IEEE Parallel and Distribute Systems pp
1245-1255, 2013.
[4] H. Mi, H. Wang, G. Yin, H. Cai, Q. Zhou,
and T. Sun, “Performance Problems
Diagnosis in Cloud Computing Systems by
Mining Request Trace Logs,” Proc. IEEE
Network Operations and Management Symp.
(NOMS), pp. 893-899, 2012.
[5] P. Barham, A. Donnelly, R. Isaacs, and R.
Mortier, “Using Magpie for Request
Extraction and Workload Modelling,” Proc.
USENIX Sixth Conf. Symp. Operating
Systems Design and Implementation (OSDI),
pp. 259-272, 2004.
AUTHORS
First Author
Second Author
325
INTERNATIONAL CONFERENCE ON CURRENT INNOVATIONS IN ENGINEERING AND TECHNOLOGY
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
ISBN: 378 - 26 - 138420 - 5
www.iaetsd.in