Networked Storage-SAN Expert for Implementation Engineers

There are 12 Solaris hosts connected to two fabrics and a DMX2000. All the hosts are dual-pathed using Emulex HBAs. All 12 HBAs in each fabric are connected and zoned to one FA port. Each host is doing 2000 IOPS. Each I/O size is 64KB.

Two of the hosts are receiving the following error in their /var/adm/ messages file: SCSI transport failed: reason 'timeout': retrying command What do you advise the customer to do?

The customer has the following configuration: ControlCenter recently upgraded to 5.2 SP1 SAN Manager 5.2 Connectrix Manager 7.01 FibreZoneBridge is at 2.2.6. The customer is migrating from an NT Service Processor to a 1U Server. They installed Connectrix Manager 8.01 and reinstalling FibreZoneBridge 2.2.6 using all the defaults in the setup program. This upgrade was successful but SAN Manager can not discover the switches.

What is the issue?

Customer added three new servers to an existing SAN consisting of two fabrics. Each fabric consists of two switches with two ISLs. This change negatively impacted performance. Throughput is as follows: 20MB/s on Host HBAs at 4K block size 50MB/s on Symmetrix on the FA ports 60MB/s on each ISL port The physical disks are 80% utilized but with 100% cache-hit ratio.

What should the customer do to improve performance?

Customer requires static SAN routing to manage the multiple links between their SANs.

Which Connectrix B-series policy allows static routing?

You are merging MDS, B-series and M-series switches. The M=series switch merged with the B-series switch. The B-series switch merged with the MDS switch. The MDS switch would not merge with the M- series switch.

What is the problem?

A customer has configured iSCSI devices on a Windows using DHCP. The DHCP lease expired and now the customer can not see the iSCSI devices. The Windows host is using the Microsoft iSCSI driver version 1.0.6 and is connected to an M-series 1620 running firmware 4.4.0.

How would you fix the problem?

Click the Exhibit button.
The customer has multiple MDS 9500 and MDS 9200 directors and switches. They need to have the firmware of the MDS 9216 upgraded. The Fabric Manager's Software Install Wizard shown in the exhibit shows a problem that must be addressed before continuing the upgrade.

What is likely to be the problem?

Using EMC's Best Practices, what is the maximum number of switches for a full mesh fabric using DS- 16B3 switches?

Click the Exhibit button. In the diagram the iSCSI servers can see the CLARiiON Array but not the DMX.
Local fiber attached servers can see all adapters in both the CLARiiON and DMX arrays.

What is the most likely cause of this issue?

The customer is preparing the B-series SAN environment to eliminate the presence of edge SAN devices.
The administrator has manipulated the Core-PID format on all edge switches to be of unique value.
Without realizing the impact of change, the attached HP-UX servers have subsequently lost access to their SAN devices.

What has happened?