COD
DLPAR
One big LPAR
Two separate p690 machines
第1题:
A customer has a 16-core Power 570 with 64 GB RAM. They have 2 VIO servers and 4 LPARs, all running very heavy database applications. Storage is virtualized. The storage is SAN based and configured via MultiPath IO on the LPARs. The users are complaining about poor performance. The VIO servers are reporting low CPU load and little memory usage. The LPARs show 100% disk busy time and are running very slowly, approximately 6-10MB/sec. The SAN reports LUN utilization of less that 20% busy time. What should be done to resolve this problem? ()
第2题:
Companycom has seven days of vmstat data on a p5 590. The server has a total of four test/dev LPARs and 4 production LPARs of one processor each. On the four test/dev LPARs, the data shows 90-95% peak utilizations with a duration of from 5 to 10 minutes. The average utilizations of all LPARs were between 15-30%. All of the applications have different times when they peak. What is the best configuration to fully utilize the processor resource?()
第3题:
A solution for two batch applications was proposed in Companycom meeting. They are currently running in two LPARs on a p690. On a daily basis, both applications use 15% of the machine for their workload, and increase to 100% when their weekly reports are run. Since the reports are not run on the same day, what is the best solution?()
第4题:
After rebooting the machine, Companycom notices that the machine is not performing as well as before. The applications are running slower, and their CPU utilization has increased. How can the customer check to see if a CPU was deallocated?()
第5题:
Companycom is planning a p5 595 running AIX 5L 5.3 LPARs. They would like to design the system using physical resources with maximum resource efficiency. One LPAR requires a peak load of 30 processors during the week with a low of 1 processor required on the weekend. The other LPAR requires only a single processor most of the time, but needs at least 24 processors at the beginning of the weekend to finish its weekly processing in the window allocated. Which of the following solutions is most appropriate to continuously maximize resource use for these two LPARs?()
第6题:
You need to design the Service Broker queues that will be used to submit requests for long-running reports. Your solution must minimize complexity of using the queues by applications. What should you do?()
第7题:
Configure MPIO load balancing on the LPARs.
Increase the SCSI queue depth per hdisk device.
Increase the number of virtual SCSI devices per LPAR by 2.
Increase the path priority on both virtual SCSI devices on the LPARs.
第8题:
Specify one queue for each client application that can submit report requests.
Specify one queue for each of the two report-generation services.
Specify one queue for all long-running report requests.
Specify eight queues:One for each service, and one for each client application.
第9题:
Modify the short cut for the Win 16 based application to enable it to use multiple processors.
Set the ‘Use multi-processor feature if available for Win 16 applications’ through the accessibility option.
Replace the Win16-based applications with available Win32-based applications.
None of the above options provides a valid solution.
第10题:
COD
DLPAR
One big LPAR
Two separate p690 machines
第11题:
Use micropartitioning for test and QA environments, and DLPAR for the production applications
Use micropartitioning for test and QA environment, and CPU pooling for the production applications
Create standard LPARs for all test, QA, and production applications to keep all partitions separate and isolated.
Create DLPAR partitions for all test, QA, and production applications to keep test from disturbing production environment.
第12题:
Run the vmstat command
Run the iostat command
Run the lparstat command
Check the logs in the /var/adm/ras directory
第13题:
Companycom is planning to do server consolidation on POWER5. The customer has a test environment, a QA environment and a production environment for two applications. Both the test and QA environments are very small and are only utilized when new releases are being developed and tested. They require less than 1 CPU each. Both production applications require 2 CPUs for day-to-day operation and can require up to an additional 2 CPUs each during abrupt peak requirements. What is the best design for the LPARs?()
第14题:
Companycom wants to deploy WebSphere in an HACMP environment, running on two servers in LPARs. These servers will also host the test and dev LPARs. The customer would like to use the test and dev resources to support an HACMP failover. What is the best way to accomplish this task?()
第15题:
A space-constrained customer is currently running AIX 5.2 on a number of p650 servers and a pair of new p5-550 servers. They have four identical new applications each requiring a minimum of seven processors and 96GB of memory. Which solution would best meet the customer’s requirements?()
第16题:
Companycom is considering a partitioned solution with one of the LPARs utilizing 12 CPUs. Multiple applications will be loaded on this large LPAR. However, the customer wants to ensure some tasks will have higher priority access to the CPU and memory resources. Which of the following solutions would best satisfy the requirement?()
第17题:
You replace the uniprocessor computer in your design department with new MPS-compliant computers. Each computer has two 550-MHz processors. You install design software that includes Win32, Win16, and DOS-based applications. When users use a Win16-based application, they do not notice an improvement in performance compared to using the uniprocessor computers. What should you do? ()
第18题:
Create a single Web site that contains both Web applications.
Create two Web sites, one for each Web application.
Assign both Web applications to the same application pool.
Assign each Web application to its own application pool.
第19题:
Set up the server in SMP mode and use WLM to manage resources effectively
Set up eight dedicated LPARs and use DLPAR to manage resources effectively
Set up a shared pool with eight capped LPARs including production and test/dev
Set up a shared pool with eight uncapped LPARs with priority weight to Production
第20题:
Rollback the transaction in both applications.
Rollback the transaction in one of the two applications.
Decrease the lock timeout value for both applications, thereby causing a lock timeout situation.
Send a message to let the administrator know of the situation and decide which application or applications should be terminated.
第21题:
Loadleveler
AIX Workload Manager
Partition Load Manager
Tivoli Provisioning Manager
第22题:
Dynamic LPAR
Workload Manager
Partition Load Manager
Shared Processor Pools
第23题:
Run the test and dev LPARs in an uncapped mode
Activate COD resources if a failover should occur
Use DLPAR to move resources between the test and dev and the HACMP LPAR
Stop the test and dev environments, bring up new profile for HACMP failover
第24题:
a list of existing network applications currently in use on the network
network audit results to uncover any potential security holes
a planned Layer 2 design solution
a proof-of-concept plan
device configuration templates