Are there limitations on the number of disks for different operating systems?
Windows and Linux (e.g., Redhat and SUSE) have basic limitations on the number of disks.

Windows
http://support.microsoft.com/kb/310072
The maximum number of disks is 8 (buses per adapter) multiplied by 128 (target IDs per bus) multiplied by 255 (LUNs per target ID) or 262144 per adapter.

Linux
http://www.linuxjournal.com/article/6238
There is a hard limit of 16 disks per major block number, and there are eight major block numbers for a SCSI controller. This limits the number of disks the operating system can describe to 128 disks per SCSI controller.

Linux
http://www.linuxjournal.com/article/6238

Who provides MPIO drivers for operating systems such as Linux (especially Red Hat and Debian) and UNIX?
For Linux and VMware the MPIO is part of the operating system.

What is a SAN?
A storage area network (SAN) is a high-speed special purpose network that interconnects different kinds of data storage devices via Fibre Channels with associated data servers on behalf of a network of users. A SAN can be clustered in close proximity to other computing resources, but may extend to remote locations for backup and archival storage using wide area network carrier technologies.

SANs support disk mirroring, backup and restore, archival and retrieval of archived data, data migration from one storage device to another, and the sharing of data among different servers in a network. SANs can incorporate subnetworks with network-attached storage (NAS) systems.

What is an arbitrated loop?
A Fibre Channel topology that defines the connections between hosts and devices. Two or more ports can be interconnected, but only two at a time can communicate. The loop supports up to 126 devices and one fabric attachment on one loop.

What is a WWN?
A World Wide Name (WWN) is a unique 64-bit address assigned to a Fibre Channel device during manufacturing.

What is an SBOD?
A switched JBOD.

What is the “Multipath Agent” on the License Management screen?
MPA101 (Agent) is part of the SMS Suite for Multipath I/O (MPIO).

What is “Multipath I/O” on the License Management screen?
Multipath I/O (MPIO) is the software that enables Failover and Path Management.

Does the RRS feature use a write through or write back method (i.e., when is the write data acknowledged or synchronized with the host)?
The Remote Replication – Synchronous (RRS) feature performs synchronous writes to the destination drive before acknowledging the writes to the host (i.e., it operates in write through mode). When the host gets an acknowledgment a second copy of the data is already on a remote drive. However, if a problem occurs with the connection and replication cannot proceed, the RRS will fall back to a “logging” mode wherein it will buffer the data to be replicated. When the connection is restored, the data will be pushed to the remote site again.

Is there a way to manually synchronize a Remote Replication – Synchronous (RRS) disk?
The -rejoin option restarts a replication session that may have been stopped earlier as a result of link connectivity issues or manual suspension. The syntax for the command is:

After a raidset rebuild due to a drive failure, do I use the replace command to move data to the replacement disk?
Using the replace command will start a second rebuild to the replacement disk. The recommended procedure is to use the exchange command feature accessible from the System Manger GUI. This will set the rebuilt drive to be a primary drive on the raidset, and the new replacement drive to be part of the spare drives.

How are SBODs named??
SBODs are assigned a name automatically when they are attached to an iQ2850 or iQ2880T system. When the storage controllers boot up, a unique name is assigned to each attached SBOD by the controllers.

The controllers sort the enclosure IDs (WWNs) of the SBODs to create a naming sequence. The first half of an SBOD name comes from the first eight characters of the name assigned to the iQ2850 or iQ2880T system. The second half of the SBOD name is made up of “SBOD” and a unique number.

With storage controller firmware revision 6140, the unique number was simply the sequence number in the sort beginning with “1.” Storage controller revision 6150 modified this to be two digits, the first of which is the loop number (0 or 1), and the second is the sequence number within the loop.

How do I check the revision of FC-SATA interposer firmware?
To identify the current interposer firmware revision, type satarev, all at the command-line interface.

How do I delete a raid array that was created for testing purposes?
To delete a raid array you must first delete all vdisks and then delete the pool that was created on the raid array. Once the vdisks and pool are removed, the raid array can be deleted.

What is the maximum number of telnet sessions supported by an iQ2880T?
The iQ2880T supports a maximum of five concurrent telnet sessions. On the sixth connection, the following message will be displayed:

Telnet session rejected
Maximum clients limit reached

What are best practices for obtaining the optimal performance on an Exchange database with iQstor storage systems?
First, separate the transaction log files and database files which will be stored on different pools/disks. Use a Raid 1+0 for transaction logs and a Raid 5 for database files. The optimal Raid 5 performance is obtained by using five to eight disks per raidset, with one pool and one vdisk in each pool. Next, distribute the load across controllers and host ports:

• Do not assign all vdisks to a single host port.
• Do not assign too many vdisk per pool. (one per pool is optimal)
• Do not enable zeroing out of deleted Exchange database pages since this will almost double the number of I/Os to the database.

How do I resolve the issue that a disk device can be found in the device manager, but it cannot be found in the disk manager?
This issue may arise if another version of a Multi-path device exists on the operating system platform and is not removed prior to installing iQstor’s MPIO or if the MPIO installation is incomplete. To correct this issue, uninstall any alternative Multi-path device and iQstor’s MPIO, and then reinstall iQstor’s MPIO.

How are MPIO disks created?
For MPIO 50xx series, all SCSI disks will be created into MPIO disks if their VendorID and ProductID match with the register parameter:

HKLM\SYSTEM\CurrentControlSet\Control\MPDEV\MPIOSupportedDeviceList

For MPIO 51xx series, SCSI disks are not created into MPIO disks by default. The user can create MPIO disks with the SAN manager if there is an MPIO license.

Which versions of the MultiPath package include a signed file from Microsoft?
All MPIO solutions (UMPD, SMS 5100, SMS 5200) are signed by Microsoft.

What is the maximum number of vdisks/LUNs per pool?
The maximum number of vdisks per system is 508 for the iQ2880T and 504 for the iQ2850. The vdisk/LUNs limit is not tied to the pool or number of pools. You can create the vdisks/LUNs up to the limit from the same pool or from multiple pools.

What is the operating temperature range for the iQ2880T and the J2880?
The iQ2880T and the J2880 are designed to operate from 10 degrees Celsius to 40 degrees Celsius ambient temperature.

What is the power consumption for the iQ2880T and the J2880?
The measured power consumption for an iQ2880T configured with 15 x 500GB SATA (Model ST3500630NS) HDDs, 2x SP288 controller modules, 4G memory and a EM28 with BT20 battery module is:

At power or start up: 656 Watts at 6.13A (surge)
In idle or stand by: 345 Watts at 2.96A
Running test software (access): 385 Watts at 3.31A
Idle (empty chassis, no HDDs): 149 Watts at 1.29A
HVAC information: 385Watts (at 3.31A) = 1,314.005 BTU/hr

At power or start up: 4.36A (surge)
In idle or stand by: 2.88A
Running test software (access): 3.14A
HVAC information: 364Watts (at 3.14 A) = 1,242.33 BTU/hr

At power or start up: 534 Watt at 5.26A (surge)
In idle or stand by: 237 Watts at 2.39A
In idle or stand by: (empty chassis, no HDDs): 81 Watts at 0.75A

How does the replication function work using iSCSI?
Replication modes can be either synchronous or asynchronous. In the synchronous mode (RRS), the acknowledgement is not sent to the host until the data is successfully written to the remote unit. In this mode, the write-order is guaranteed. In the asynchronous mode (RRA) the acknowledgement is sent even before the data is successfully written into the remote site. In fact, the data can be sent to the remote site much later, pursuant to the replication schedule configured by the SAN administrator. This mode utilizes the latency on long distance. The iQ2850 supports only RRA and the iQ2880T supports only RRS.

What is the difference between a Fibre Channel storage network and an iSCSI storage network?
Fibre Channel (FC) and iSCSI are network protocols that enable SAN fabrics. While both types of networks can coexist, they are generally directed at different end-user markets. With its low latency and high performance and reliability, Fibre Channel is the dominant SAN fabric implemented in the data centers of mid and enterprise companies. iSCSI storage networks are completely distinct from FC networks in that the underlying protocol is based on TCP/IP. Since TCP/IP is more accessible and more affordable, iSCSI is the choice for small and mid-sized businesses that want to use their existing infrastructure and cannot afford the cost of dedicated storage administrators. Another popular use for iSCSI is in remote replication for disaster recovery applications since iSCSI has a much longer reach than Fibre Channel (with its limitation of approximately 6miles/10km).

How is the iQ2850 different from the iQ2880T?
Both of the systems are similar in that they have identical architecture and back-end design. They differ in the front end in that the iQ2850 has four Gigabit Ethernet ports with TCP/IP offload engines while the iQ2880T has two 4Gb/s FC ports per controller for host connectivity.

Can more than one server access data storage on iSCSI based SANs?
Yes, this is one of the more attractive attributes of SANs as a whole and one of the primary benefits of iSCSI specifically. Fibre Channel SANs have made file sharing by multiple servers a reality, but not all SMBs can afford the expense of Fibre Channel SANs. iSCSI allows SMBs to share files by multiple servers at an affordable price.

Why is a TOE required in iSCSI solutions?
An iSCSI TOE offloads the TCP/IP protocol processing from the server CPU to the host bus adapter processor. iSCSI host bus adapters or TOEs are required when the additional performance enhancement justifies the additional expense of using a host bus adapter for iSCSI rather than using a software-based iSCSI client (initiator).