Deck 1: Hitachi Vantara Certified Expert Performance architect

Question 1

Your VSP G1000 customer has an HDP pool with 600 GB drives and they need more IOPS in the pool. Which action will satisfy the requirement?

A) Spread the pool over storage tiers.
B) Add 1.2 TB drives to the pool.
C) Add more 600 Gb drives of the same type.
D) Create a new HDP pool.

Answer 1

Spread the pool over storage tiers.

Question 2

You are sizing a storage system for a response-sensitive database application in a VMware environment. Which two statements are true in this scenario? (Choose two.)

A) The cache hit percentage is determined primarily by the amount of cache provided by the storage system.
B) The peak load requirements of the database and filesystem is a key factor in determining the resources required to meet response time requirements.
C) The server virtualization will change the I/O profile of the application as viewed by the storage.
D) Sizing estimations for a database workload should be based on a steady state and then an extra 10% resources for headroom should be added.

Answer 2

The peak load requirements of the database and filesystem is a key factor in determining the resources required to meet response time requirements.
The server virtualization will change the I/O profile of the application as viewed by the storage.

Question 3

A VSP G800 HDT pool contains FMDs, NL-SAS drives, external storage with 15K 600 GB SAS drives, and external storage with 10K 900 GB SAS drives. Which tier level contains the NL-SAS drives in this configuration?

A) 1
B) 2
C) 3
D) 4

Answer 3

1

Question 4

Immediately after migrating to a new HDT pool, Hitachi Tuning Manager shows that Tier 1 SSD, Tier 2 SAS, and Tier 3 NL-SAS are not busy for performance but Tier 1 and Tier 2 are over utilized for capacity. From this analysis, which action should be performed?

A) Re-asses after HDT performs page migration cycles.
B) Add more drives to Tier 1.
C) Add more drives to Tier 1 and Tier 2.
D) Remove the SSD tier.

Question 5

Response times in an HDP pool have been increasing consistently with each additional workload during a large scale migration. The response times may become unacceptable with subsequent migrations. The purchase of additional equipment is an acceptable solution. What are two valid actions in this scenario? (Choose two.)

A) Increase the pool free space and alert thresholds to the maximum permitted levels.
B) Expand the pool with volumes from new parity groups to reduce utilization levels.
C) Apply the SOM to provide tier property information that assesses suitability for HDT and deployment of flash.
D) Use more, smaller, pool LDEVs per parity group to improve the efficiency and utilization of the queues.

Question 6

Which three statements are true regarding the Hitachi external storage architecture? (Choose three.)

A) Enabling Cache Mode increases exposure to problems with high Cache Write Pending.
B) Enabling Cache Mode allows destaging asynchronously from host I/O.
C) Enabling Cache Mode always improves performance.
D) A System Option mode can force in sequence destaging even when destaging asynchronously.
E) Disabling Cache Mode for an external LDEV eliminates the use of cache.

Question 7

A user with a diskless VSP G800 uses ShadowImage from the G800 with a virtualized VSP G200 system. Application performance suffers during the frequent ShadowImage resynchronization operations. What should be done to reduce the impact of the ShadowImage throughput in this scenario?

A) Configure both sets of disks into an HDT pool and let the HDT scheme manage the different workloads.
B) Define the ShadowImage target volumes to use dedicated external FC paths using dedicated RAID groups.
C) Set Copy to 15.
D) Increase the VSP G800 cache size to improve the application cache hit rates.

Question 8

A customer has an HDP pool with four array groups and 600 GB SAS drives operating at 85% utilization. Four array groups of the same configuration were added to the pool. In this scenario, which metric should be reviewed to validate optimal pool performance?

A) pool total IOPS
B) pool I/O profile
C) pool array group utilization
D) pool read/write ratio

Question 9

A customer has several Oracle databases with different workload profiles residing on old HUS systems. The customer wants to consolidate workloads on a new VSP G800 system to increase database performance and to reduce costs. Which configuration would you use to accomplish this task?

A) Put all of the databases in a single HDT pool.
B) Put each database in its own HDT pool.
C) Put databases with the same workload profile together in their own HDT pool.
D) Put all database tablespaces and indexes in the first HDT pool, all redo logs in the second HDT pool, and all undo logs in a third HDT pool.

Question 10

Your customer has an HDP pool dedicated to interactive production workloads. Which two applications are good candidates to share this HDP pool? (Choose two.)

A) customer relationship management
B) customer service call taking and dispatch
C) software development  
D) backup repository  

Question 11

A customer has a VSP G800 storage system with an HDT pool with internal SAS drives and a virtualized third-party array using SATA RAID 6. The virtualized array is connected to the VSP G800 using four 8 Gbps FC paths. The customer reports performance problems on an OLTP application deployed on a virtualized server using the HDT pool. What are two reasons for this problem? (Choose two.)

A) The workload includes significant random writes to SATA over RAID 6.
B) The tier relocation is not assisting the OLTP application.
C) The size of Tier 1 is not sufficient to accumulate all of the hot pages.
D) The speed of the FC links is insufficient.

Question 12

When testing an external LUN with default mapping on two paths, performance appears constrained and the utilization of the RAID group in the external storage system never exceeds 35% busy no matter how many test threads are generated. You want to increase the utilization of the external RAID group. Which two actions will accomplish this task? (Choose two.)

A) Increase the external path LUN queue depth setting.
B) Increase the number of LUNs mapped from the external RAID group.
C) Change Cache Mode to "Disabled".
D) Increase the server HBA LUN queue depth to 64.

Question 13

A VSP G800 customer has one HDT pool with three tiers and they report a high frequency of relocations up and down between Tier 1 and Tier 2, affecting the HDT pool performance. Which two actions will help solve the issue? (Choose two.)

A) Decrease the monitoring period of the HDT pool.
B) Increase the size of the grey zone between Tier 1 and Tier 2.
C) Change the HDT pool setting from Period mode to Continuous mode.
D) Increase the size of Tier 2.

Question 14

You are deploying an external storage system and its connection to a VSP G1000. The VSP G1000 has four external pools. You must ensure balanced utilization of the external storage system parity groups within each pool. What would you do to satisfy this requirement?

A) Use only one external volume per external parity group with the default UVM queue depth.
B) Slice each external parity group to provide a pool volume for each pool.
C) Ensure that volumes from a parity group in an external system are exclusively associated with the same external parity group in the storage system connected to the host.
D) Provide eight paths for each volume presented by the external storage system to the storage system connected to the host.

Question 15

Relocation cycles are regulatory failing to complete in an HDT implementation where the customer is also reporting lower than expected performance. What should be done in this scenario?

A) Change to Continuous mode and increase the monitor period cycle time.
B) Change to Period mode and increase the monitor period cycle time.
C) Change to Period mode and reduce the monitor period cycle time.
D) Change to High-Performance mode and reduce the monitor period cycle time.

Question 16

What are two benefits of placing transaction log files on RAID 1+0 disks for Microsoft SQL Server databases rather than RAID 5 disks? (Choose two.)

A) Multiple transaction log files are supported.
B) Faster recovery after disk hardware failure is realized.
C) Capacity utilization rates are improved.
D) Microsoft recommendations are followed.

Question 17

A customer plans to migrate an interactive application from a VSP to VSP G1000. The application has consistent access across all its capacity with low hit rates and no particularly high I/O access areas. The application latency must be less or equal to 1.5 ms. Which configuration do you recommend to satisfy the requirements?

A) an HDT pool consisting of SAS 10K and FMD drives
B) an HDP pool with FMD drives
C) an HDP pool consisting of NL-SAS and FMD drives
D) an HDP pool consisting of external LDEVs provisioned on a VSP F400 system

Question 18

You are planning for a performance upgrade for a transaction application. Which three metrics are necessary to determine the required back-end storage configuration in this scenario? (Choose three.)

A) cache read hit ratio
B) read/write ratio
C) transfer rates
D) queue depth
E) IOPS

Question 19

Which two statements are true regarding HDP pools? (Choose two.)

A) Using the largest possible pools is a recommended practice.
B) Exposure to failure and recovery from failure must be considered when recommending the maximum size of a pool.
C) HDP pools can be used for sizing HDT pools.
D) Tier migration is based on performance and space capacity.

Question 20

A customer with a VSP G1000 is using HDT with SAS 15K drives and external storage. The customer wants the daytime response-sensitive workload to benefit from tiering. They also want to minimize the impact of relocations on the daytime operation itself. In this scenario, which action will satisfy the requirements?

A) Enable monitoring only during the daytime hours (9:00 AM to 5:00 PM) and only enable relocation at other times.
B) Disable monitoring during daytime hours (9:00 AM to 5:00 PM)
C) Enable active flash so that all relocations can be completed in short bursts.
D) Use continuous mode with a 30-minute cycle time and define relocation during daytime hours (9:00 AM to 5:00 PM) with a low relocation speed.

Question 21

A customer complains about high response times. This occurs frequently with bursts of write I/O at 6,000 IOPS for 10 second periods during each minute. The write response is 60 ms during these 10 second periods. The average IOPS rate reported by Hitachi Tuning Manager over 1 minute is 600. In this scenario, what is the average response time over 1 minute?

A) 10 ms
B) 30 ms
C) 60 ms
D) 90 ms

Question 22

How does the HDLM Extended Round Robin algorithm impact sequential I/O?

A) It continues to use the same path for a user-specified count of sequential I/O requests.
B) It remembers a user-determined number of addresses to better implement sequence detection for the storage.
C) It buffers and reorders I/O requests to improve sequence detection by the storage.
D) It makes a path decision for every I/O request.

Question 23

For which two purposes is the SVP Performance Monitor suitable? (Choose two.)

A) Reporting by device owner
B) Annual trend analysis
C) Real time observations
D) A reference for verifying the accuracy of other performance tools

Question 24

Which two conditions will cause high Cache Write Pending on a VSP G8000? (Choose two.)

A) over-utilized external storage
B) high port utilization
C) inadequate WAN bandwidth for TrueCopy Asynchronous
D) attempting to write to a RAID group at a rate beyond its ability

Question 25

Which two conditions are amplified by the use of longer data intervals for analysts? (Choose two.)

A) There will be more arbitrary data loss when viewing on a typical computer screen.
B) Analysis tools will require more processing time and computing resources.
C) The peaks values reported will reflect interactive user experience less accurately.
D) Analysis tools will require less processing time and computing resources.

Question 26

A large bank is completing a capacity planning exercise and wants to upgrade their current VSP G1000 to properly support their online banking application for the next 18 months. The IT department wants to know where they should invest to achieve the transactional Service Level Objective (SLO). In this scenario, which three recommendations would you make? (Choose three.)

A) Measure throughput of all SAN directors.
B) Export two months of Performance Monitor data.
C) Measure the storage resource utilizations.
D) Request a statement of growth requirements.
E) Measure the existing I/O profile.

Question 27

In a database environment, the system administrator complains that high write response times are triggering frequent automated alerts. However, utilization for processor, paths, and array groups is low. Database users are not complaining. What causing this problem?

A) The RAID level is inappropriate.
B) The storage system has too few processors.
C) The storage media devices are too slow.
D) Asynchronous destaging of writes occurs in bursts.

Question 28

A VSP G1000 performance analysis shows that server response times for virtualized volumes are high, 100 ms and higher, but HDD utilization rates in an external HUS 150 never exceed 35%. There are no processor or path constraints in the configuration. The workload is steady and stable. Which two actions will optimize this configuration? (Choose two.)

A) Increase the external path LUN queue depth in the VSP G1000.
B) Spread the server load across more and smaller LUNS per RAID group in the HUS.
C) Break up each HUS LUN into several smaller volumes in the VSP G1000.
D) Migrate the HUS LUNs to RAID groups with faster HDDs.

Question 29

A VSP G1000 customer is concerned about high VSD utilization alerts during their nightly backup and application maintenance window. To address their concerns, you want to inform them about the maximum acceptable utilization level range. What would be the range in this scenario?

A) 30-40%
B) 40-50%
C) 50-60%
D) 80% or higher

Question 30

You are recommending RAID 6 to a customer for a non-database HDT pool. What are two reasons for this recommendation? (Choose two.)

A) When capacity is pooled; the whole pool is affected when one parity group fails.
B) RAID 6 provides the most efficient use of system processor resources.
C) RAID 6 minimizes exposure to double disk failure.
D) RAID 6 provides the optimum overall performance of all RAID types.

Question 31

DP-Vol LUNs are taken from an HDP pool deployed with 128 disks (16x4D+4D RAID groups). Access is random read. Which combination of FC paths, LUN Queue Depth, and number of DP-Vols will result in a per-disk queue of 4 within this pool?

A) 2 paths, a LUN Queue Depth of 16, and 32 DP-Vols
B) 2 paths, a LUN Queue Depth of 8, and 32 DP-Vols
C) 4 paths, a LUN Queue Depth of 8, and 32 DP-Vols
D) 4 paths, a LUN Queue Depth of 16, and 32 DP-Vols

Question 32

For a VSP G1000 system serving a mission critical interactive workload, which two statements are true about VSD failover? (Choose two.)

A) The VSD receiving the workload of a failed VSD is not pre-determined.
B) Providing reserved capacity to accommodate failure is a requirement when planning for capacity for normal operations.
C) The workload of a failed VSD board always transfers to its redundant partner VSD board.
D) The workload of a failed VSD board is dynamically redistributed among surviving VSD boards.

Question 33

A high level of writes to external storage is backing up into a VSP G1000 cache, creating a high write pending problem that is affecting other workloads. Which three actions will help to minimize this problem? (Choose three.)

A) Add more cache to the enterprise storage system.
B) Add more cache in the external storage.
C) Quarantine the LUNs in a separate cache partition.
D) Set Cache Mode to "Disabled" for the external LUNs.
E) Address the throughput constraints in the external storage.

Question 34

Which three statements are true regarding a VSP G400 configuration for a Microsoft Exchange Server environment? (Choose three.)

A) RAID 1+0 is a recommended RAID level for the database volumes.
B) RAID 6 is a recommended RAID level for the database volumes.
C) HDP is a recommended configuration for database volumes.
D) Parity group concatenation is recommended.
E) Microsoft JetStress testing should be used to validate the required performance.