Storage Boards Cleanup
To make it easier to find information about HPE Storage products and solutions, we are doing spring cleaning. This includes consolidation of some older boards, and a simpler structure that more accurately reflects how people use HPE Storage.
Tape Libraries and Drives
cancel
Showing results for 
Search instead for 
Did you mean: 

MSL6000, four Drivers and Two NSR e1200-160

MSL6000, four Drivers and Two NSR e1200-160

i have MSL6060 with 2 NSR's ,four LTO3's and one controller , any body can help me to identify the scsi map between NSRs,Drives and Controller ,,, i tried to connect each NSR to two drives and terminate one drive bus ,and connect a scsi bus for one drive with the controller in the first "UPPER" group ,,, and connect the the second NSR to the other two drives with two terminators , one on each drive ,,,, connect each NSR to different Fabric switch ,,,

this solution is not fully correct , because i can only see one NSR and two drives on the cell manager server ,,, through mapping all ports had been identified auto assigned on both NSR's , i can get all the ports that is discovered on the fabric switches on the MSL web page ,,, Just help me to identify the missing sittings ,,,

Thank You
Ayman Jweinat
5 REPLIES
rich pattison
Trusted Contributor

Re: MSL6000, four Drivers and Two NSR e1200-160

From your diagram it looks like the connections from NSR's to drives are correct (1 drive per SCSI bus). If the server cannot "see" the drives then you need to make sure there is a data path from the server to each of the NSR's. Using the auto-assigned map should ensure that all drives will be mapped to all hosts.

On the mapping page of each NSR make sure the WWID of the server HBA is listed - if not - the server has no link to the NSR. Also check the status of the fibrechannel port from the "modules" page - it should be set as an N_port/Fabric state "UP".

It' probably worth checking the Switch to make sure the NSR ports are logged-in OK.

This sounds more like a SAN connectivity problem the the NSR/Library.

Re: MSL6000, four Drivers and Two NSR e1200-160

Dear
the mapping for MSL is set to auto assigned for all hosts WWN's, the fabric status for NSL is up , i have two cisco mds switches that connected to two NSR's as recommended , i can only discover the changer and two drives which are connected to the fabric related to the controller, i test to change the fiber switching or to disconnect one switch over the solution, i have HBAAnyWare software which help me to identify the connectivity, the main idea is : i cannot discover the NSR and the two drives which are away from the MSL controller so in the device manager i still have only one changer with two drives ????
Marino Meloni_1
Honored Contributor

Re: MSL6000, four Drivers and Two NSR e1200-160


- first, the cabling is not fully correct, the drive that is daisy chained with the library controller should have the terminator, so connect nsr port >> Lib Controller >> Drive Canister >> terminator
- Then, each NSR is independent, each NSR should be configured, and map assigned to hosts. You should be sure about your zoning also. The host should see both NSR, so it can see both maps

Marino

Re: MSL6000, four Drivers and Two NSR e1200-160

Dear
can you draw the suggested diagram from you,i feel my idea mismatched with yours ,,, i wish you to send a me a diagram in order to implement it ...

Thank You...
Marino Meloni_1
Honored Contributor

Re: MSL6000, four Drivers and Two NSR e1200-160

here is the draw related to cabling.
If you cannot see the drive in the server, you should follow this steps:

check in each NSR if all devices appear and are up on the scsi channel

create a map in each NSR with all the components
(remember that for W2K3 servers, a LUN 0 is mandatory, or the server will not check if there are devices on that bus)(check this on each NSR)
present these maps to each server (to be made on both NSR)
Check that each NSR are visible by the servers (check zoning)

be sure to do a rescan in device manager after you cahnge the maps in nsr, or if the rescan is not enough, you may have to reboot your server