Academia.eduAcademia.edu

Figure 5 – uploaded by Fernando Joaquim

See full PDFdownloadDownload figure
Since there are three nodes available it seems obvious each node to get a job. But actually job 232 could have chosen a slot from the fastest machine since the normal priority Submitting Jobs Initially 2 Low priority and 1 Normal priority jobs will be submitted. taken because the resources are firstly ranked by its availability, as described in 4.3.5.2.

Table 5 Since there are three nodes available it seems obvious each node to get a job. But actually job 232 could have chosen a slot from the fastest machine since the normal priority Submitting Jobs Initially 2 Low priority and 1 Normal priority jobs will be submitted. taken because the resources are firstly ranked by its availability, as described in 4.3.5.2.

Related Figures (49)

Figure 1.1: The world’s first web server 1.1.2 Current Projects The current large scale project at CERN is the very well known Large Hadron Collider LHC) [4]. Large because of its 27 Km (using the tunnel from LEP) and Hadron because protons and ions are to be collided. It will be used to recreate the conditions just af- ter the Big Bang by colliding two particle beams at very high energy (about 7TeV per proton) which makes them to travel at more than 99.9% of the speed of the light. The project started back to the 1980s and in December 1994 the Cl construction, at a total cost of about 6 billion CHF. ERN council approved its The current large scale project at CERN is the very well known Large Hadron Collider
Figure 1.1: The world’s first web server 1.1.2 Current Projects The current large scale project at CERN is the very well known Large Hadron Collider LHC) [4]. Large because of its 27 Km (using the tunnel from LEP) and Hadron because protons and ions are to be collided. It will be used to recreate the conditions just af- ter the Big Bang by colliding two particle beams at very high energy (about 7TeV per proton) which makes them to travel at more than 99.9% of the speed of the light. The project started back to the 1980s and in December 1994 the Cl construction, at a total cost of about 6 billion CHF. ERN council approved its The current large scale project at CERN is the very well known Large Hadron Collider
Figure 1.2: The CERN accelerator complex
Figure 1.2: The CERN accelerator complex
Table 3.1: Comparison between HTC and HPC Clusters have also evolved in different ways depending on their purpose. One can differenti- ate between High-Performance-Computing (H PC), High-Throughput-Computing (HTC), High-Availability (HA) clusters and Load-Balancing (LB) clusters. The concept between HPC and HTC is often mixed up and, actua ly, there’s always some overlap between all of these classifications. The main characteristic that distinguish between HPC and HTC is the fact of the latter being designed for bet ter global performance over a long period of time, instead of the best performance for a short period of time.
Table 3.1: Comparison between HTC and HPC Clusters have also evolved in different ways depending on their purpose. One can differenti- ate between High-Performance-Computing (H PC), High-Throughput-Computing (HTC), High-Availability (HA) clusters and Load-Balancing (LB) clusters. The concept between HPC and HTC is often mixed up and, actua ly, there’s always some overlap between all of these classifications. The main characteristic that distinguish between HPC and HTC is the fact of the latter being designed for bet ter global performance over a long period of time, instead of the best performance for a short period of time.
Figure 3.1: JMS general architecture 3.3. Comparative analysis of available JMS
Figure 3.1: JMS general architecture 3.3. Comparative analysis of available JMS
Table 3.2: HTC comparison table 3.3. Comparative analysis of available JMS
Table 3.2: HTC comparison table 3.3. Comparative analysis of available JMS
Figure 4.1: Machine roles in Condor Figure 4.1 on page 22 shows a usual configuration of a Condor system, where the
Figure 4.1: Machine roles in Condor Figure 4.1 on page 22 shows a usual configuration of a Condor system, where the
Figure 4.2: System physical architecture
Figure 4.2: System physical architecture
Figure 4.3: Condor main daemons 4.2 System specification Since the role of each computer is defined, the very next step is to install Condor and
Figure 4.3: Condor main daemons 4.2 System specification Since the role of each computer is defined, the very next step is to install Condor and
Figure 4.4: Condor job state diagram
Figure 4.4: Condor job state diagram
Figure 4.6: System state diagram 4.3. Implementation of the Condor configuration
Figure 4.6: System state diagram 4.3. Implementation of the Condor configuration
Figure 4.7: Slots in a dual core machine 4.3. Implementation of the Condor configuration 4.3.3. Implementing Condor non-blocking suspension By the default behavior, Condor does not suspend jobs, instead preempts them when
Figure 4.7: Slots in a dual core machine 4.3. Implementation of the Condor configuration 4.3.3. Implementing Condor non-blocking suspension By the default behavior, Condor does not suspend jobs, instead preempts them when
4.3.5.1 Load management 4.3. Implementation of the Condor configuration Because the simulation processes completely occupies one CPU core, it was defined that
4.3.5.1 Load management 4.3. Implementation of the Condor configuration Because the simulation processes completely occupies one CPU core, it was defined that
Figure 4.9: Desktop CPU load and interference with heavy processes load if a heavy process was running, affecting users’ processes performance.
Figure 4.9: Desktop CPU load and interference with heavy processes load if a heavy process was running, affecting users’ processes performance.
No surprises so far, the cluster is full and so job 233 will take the resource of a Low job will be submitted to the system, but let’s mark this one as Long. refresh of the condor_status, the job has already quit the execution machine.
No surprises so far, the cluster is full and so job 233 will take the resource of a Low job will be submitted to the system, but let’s mark this one as Long. refresh of the condor_status, the job has already quit the execution machine.
Figure 4.10: Load control - machine fills up with local jobs
Figure 4.10: Load control - machine fills up with local jobs
Figure 4.11: Load control - machine waits until gets free 4.4. Tests and analysis of results
Figure 4.11: Load control - machine waits until gets free 4.4. Tests and analysis of results
Figure 4.12: Load control - extensive test This test in intended to check most situations related to load management and how the
Figure 4.12: Load control - extensive test This test in intended to check most situations related to load management and how the
Figure 5.1: System’s usage profiles and application dependencies 5.1.1 Logical structure
Figure 5.1: System’s usage profiles and application dependencies 5.1.1 Logical structure
Figure 5.3: Simulation file structure files are located in the same directory, but when it comes to sub-simulations, created from the split of a simulation, it gets complicated to handle every file in the same directory:
Figure 5.3: Simulation file structure files are located in the same directory, but when it comes to sub-simulations, created from the split of a simulation, it gets complicated to handle every file in the same directory:
Figure 5.4: coflu_submit architecture The configuration file parser performs a similar task to the shared configuration file,
Figure 5.4: coflu_submit architecture The configuration file parser performs a similar task to the shared configuration file,
Figure 5.5: coflu_submit interactive mode include a python function which checks, parses and stores parameters into a list.
Figure 5.5: coflu_submit interactive mode include a python function which checks, parses and stores parameters into a list.
Figure 5.6: Coflu-Web architecture: Horizontal decomposition validator returns true for empty strings as well, so it the default is assumed.
Figure 5.6: Coflu-Web architecture: Horizontal decomposition validator returns true for empty strings as well, so it the default is assumed.
Figure 5.7: Coflu-Web: Vertical decomposition predefined existing resources, like those specified in 5.4.1 or include specific resources nc
Figure 5.7: Coflu-Web: Vertical decomposition predefined existing resources, like those specified in 5.4.1 or include specific resources nc
5.4.3 Physical architecture
5.4.3 Physical architecture
Figure 5.9: File structure and module template 5.5.1 Project structure To provide easy addition of features and modules, two strategies were applied: main
Figure 5.9: File structure and module template 5.5.1 Project structure To provide easy addition of features and modules, two strategies were applied: main
centralized. As a third point, the commands should execute on specific machines, so the load is distributed and users can freely switch between web and shell interfaces.
centralized. As a third point, the commands should execute on specific machines, so the load is distributed and users can freely switch between web and shell interfaces.
Figure 5.11: Coflu-Web: simulation configurations Tier 2 - Web-server module validation With a working client side validation one could omit server-side validation. But due to
Figure 5.11: Coflu-Web: simulation configurations Tier 2 - Web-server module validation With a working client side validation one could omit server-side validation. But due to
Figure 5.12: PEAR/HTML_AJAX in Coflu-Web 5.6. Tests and result analysis session persistence and therefore keeps loaded configurations available during application
Figure 5.12: PEAR/HTML_AJAX in Coflu-Web 5.6. Tests and result analysis session persistence and therefore keeps loaded configurations available during application
Figure 5.16: Job submission results in debugging mode
Figure 5.16: Job submission results in debugging mode
Figure 5.15: Coflu-Web - Submitting simulation
Figure 5.15: Coflu-Web - Submitting simulation
is the first invoked tool which outputs the generated seeds, followed by coflu_gensubmit
is the first invoked tool which outputs the generated seeds, followed by coflu_gensubmit
/run2 > condor_submit submit_ test 000 (234.000.0000) 03/31 14:17:30 Job submitted from host: <137.138.213.89:60745 >
/run2 > condor_submit submit_ test 000 (234.000.0000) 03/31 14:17:30 Job submitted from host: <137.138.213.89:60745 >
1 LOCAL RUNNING + SUBMITTED 1 CONDOR + SUBMITED 1 LOCAL B.2 CPU Load management
1 LOCAL RUNNING + SUBMITTED 1 CONDOR + SUBMITED 1 LOCAL B.2 CPU Load management
#444 STOPPED 1 local jobs 1 JOB RUNNING + 2 Submitted into Condor + 1 Local Started + 1 Local Stoped + 1 LocalStopped
#444 STOPPED 1 local jobs 1 JOB RUNNING + 2 Submitted into Condor + 1 Local Started + 1 Local Stoped + 1 LocalStopped
The home page shows general information and instructions on how to use the cluster. It’s
The home page shows general information and instructions on how to use the cluster. It’s
The submitssion page allows users to submit simulations to the cluster. To change
The submitssion page allows users to submit simulations to the cluster. To change
Figure D.4: Coflu-Web Administration
Figure D.4: Coflu-Web Administration
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved