You are designing a VNX for Block solution for a customer. The application is a high throughput
OLTP and will require the use of a volume manager to combine multiple LUNs across multiple
SPs. Each of the component LUNs resides on a RAID 5 (4+1) RAID Group.
Based on best practices, which volume manager stripe element size would you recommend?

An Oracle database is stored on an EMC VNX storage array and replicated to a second VNX array
at a DR location. EMC RecoverPoint is providing the replication. The customer is currently using a
single Consistency Group for the Oracle database.

What would you recommend the customer do to ensure optimal recoverability?

A customer has a mix of Windows 2003 and Windows 2008 virtualized guests deployed in their
vSphere v4 environment. They have purchased a new VNX array and are also upgrading their
ESX servers from V4.0 to V5.0.
Their VMware administrators are creating the new target LUNs and VMFS file systems for their
datastores. They have noticed that the alignment offset on the new array is not the same as their
old array, so they are concerned about disk crossing on the new array.
How would you respond to the customer?

A customer has deployed a legacy application on their VNX storage system that is configured with
NL-SAS drives. They are considering adding an application to this array. They have reviewed the
analyzer file and told you that the NL-SAS disk queue length is 8.

What advice would you give them?

Your customer has implemented FAST Cache and FAST VP in their application environment.
They have been monitoring drive performance, and have noticed that their Flash drives display
increased latency as they become full.
Which process can cause the slowdown?

A clone source LUN is 1 TiB in size. It performs 450 random 8 KiB IOPs with a 2:1 R/W ratio. As
part of ongoing testing, the clone is fractured for five minutes and then resynchronized. The testing
is performed every hour.
How much data (to the nearest GiB) is read from the source LUN as part of the resynchronization
process?

Your customer is operating a payment processing application on a Windows Server cluster with
the back-end Oracle databases running on Oracle Enterprise Linux (OEL). Both the application
and database environments have been presented storage from the customer’s VNX solution.
Your customer reports less than optimal performance for their online payment processing
application. While utilizing IOSTAT, utlbstat and utlestat, a high load on the buffer cache and I/O
system performance issues are found on the database servers.
What is the most likely problem?

A customer’s database resides on a RAID 5 Pool with SAS drives.
Which best practice should be followed to optimize Reserved LUN Pool disk performance?

Your customer is designing a Proof of Concept with SQL Server attached to a VNX. The storage
design includes seven LUNs on a RAID Group. The system is generally very heavily loaded, with
a particularly high load on the cache. This is a very carefully controlled environment and your
customer has asked for assistance in planning the layout of the LUN location on the physical
spindles.
Where in the RAID Group would you recommend placing the two heavily used LUNs if the
physical spindle is fully utilized?

A Snapshot Source LUN performs 500 random 4 KiB IOPs, with a 4:1 R/W ratio. At what rate is
chunk data written into the RLP when the session starts?