Storage Array Queries

[Raja N]

Hi All,
I have following doubts about storage controllers and will be thankful if any one there answers the same.

1. In an active passive controller,how passive controller takes over I/O requests when active controller fails? I would like to know the mechanism behind that.

2.How fabric switch identifies the initiator (Server node) and the target (storage node).

3. What is the difference between file level and block level access

Thanks in advance.

Raja

[Steven Clementi · Accepted solution]

1. Usually, it is the multipathing software that detets the failure and then starts using alternate paths to the other controller.

2. Fabric Switches do not identify SAN based devices as "targets" and "initiators". Generally speaking... hosts nodes are considered "initiators" and storage devices are considered "targets". The terms are an industry standard designation and have nothing to do with the switch.

3. I am not exactly sure how to convey the difference in words. I am sure there is good way and that someone will probably post it. ;o)


Steven

[Sivakumar MJ._1]

Ans 1. Active/passive means that one controller is active to process I/O requests, while the other is idle in standby mode ready to take over I/O activity should the active primary controller fail or be taken offline. An active/active configuration enables both controller nodes to process I/Os, and provide a standby capability for the other.

Ans 2. Fabric communicates using wwn numbers from Server HBA Cards and Storage Ports. Go through the SAN.

Go through this link all you questions should be clarified..

http://h20000.www2.hp.com/bc/docs/support/SupportManual/c00403562/c00403562.pdf

[Uwe Zessin]

> Active/passive means that one controller is active to process I/O requests

True for an MSA, but on an active/passive EVA both controllers can process I/Os (for different virtual disks).


There is no unique definition of what is an active/passive or an active/active array. For example, an MSA2012 or a Dell/EqualLogic PS-Series is called an active/active array, although the second controller does not present any SCSI LUNs for volumes of the first controller.

But they are called A/A, because it is possible to push down I/Os all existing paths - some that is not possible on an A/P MSA1000 series or EVA-3000/-5000.


In A/P and A/A arrays, both controllers communicate to each other via an interlink. If one controllers assumes that its partner as failed, it takes over management responsibility of the partner's ressources (logical/virtual disks, volumes, storage sets, LUNs - whatever terminology that vendor has invented ;-)

[Raja N]

Thanks for every one who put their time and knowledge for answer my question.

Can you tell me what is the difference between file level and block level access.

[Sheldon Smith]

file level vs. block level? A quick search via Google came up with the following link (along with several million others):

http://searchstorage.techtarget.com/expert/KnowledgebaseAnswer/0,289625,sid5_cid607313,00.html

[krusty]

It appears everyone has answered Q1 & 2 pretty well, so I'll focus of #3.

File Level - Typically known are NAS. This is where storage is presented to the host as a directory or drive letter and utilize such protocols as CIFS and NFS. Host was no control of the "backend" of the files (the layout). I/O is performed via the LAN (TCP/IP).

Block Level - Storage is presented ot the host as blocks. The host OS then utilizes it volume manangement software to utilize this. This allows flexibility and tuning to maximim performance. I/O is proceessed via an HBA.

Cheers,

Curt

[Sivakumar MJ._1]

NAS vs SAN....