Difference between Lun Mapping & Lun Masking.

when you map a device to front end adapter ..you assign your lun an address. When you mask your lun ..you actually make it available to the host for access.

In good old "copper" SCSI days, mapping a device on a port of the storage was the act that let the host see the device since each port of the storage was connected to a single host with a copper cable. So for historical reasons we still "map" the devices on the ports of the storage. The real reason to map devices is that the code inside the box apply strict security check to avoid unallowed access to data so "mapping" is the way to choose which processor/port can see a given device.

But with FibreChannel, you can have different hosts sharing the same port of the storage. This can be a problem IF two hosts belong to different customers OR with some operating systems that tend to put their signature here and there OR sysadmins are not willing to share their disks with someone else (these are only silly examples)

So FibreChannel required the concept of masking I.E. a filter (something like a firewall) that allowed someone to reach some devices while forbidding everything else to access the same device just llike the code does inside the box.

Things are slightly but constantly changing and with recent codes (and recent Solution Enablers) you are allowed to use what we call DLU (Dynamic Lun Addressing) that still requires you to "map" devices on the ports of the storage but allows you a fine tuning of the lun numbers used for each and every WWN.

So when you "map" a device on the frontend you allow the port (the processor) of the storage to "see" the device inside the storage, while when you "mask" the device to a given WWN you allow this specific host to see the device through a given port that already can see the same device.

I came from "3rd party" environment where we dont do masking - we just assign a LUN to a specific host hba WWN and no other host hba (with different WWN) can see that specific LUN. Needless to say the whole idea of mapping and masking confused me. Del Corno;s 2007 explanation is the first time I finally get the cause and effect connected" (zoned) to a specific storage port. 

1.first, you map 10 logical devices to the storage port.

2.then you perform masking so that host1 can see 3 of the logical devices, and host2 can see another 3 of the logical devices, and host3 can see the remaining 4 logical devices, and so on.

Hi USShivas,

There is simliar conversation in thread, https://community.emc.com/message/734348#734348. I think the replys help me a lot to understanding the Mapping and Masking in Symmetrix Array. I'd like to suggest you to go through those reply. You will understand what's the differences.

-Fenglin