On Jan 8, 2015, at 12:06 PM, Gupta, Amit amit.gu...@optum.com
wrote:
That is starting to make sense.. Let me summarize so I have got it right
Current
mkfs.lustre --fsname=temp --mgs --mdt --index=0 /dev/mapper/mpathcp1 - On
MDS/MGS
mkfs.lustre --ost --fsname=client1 --reformat
Thanks Rick and Joe for your suggestions.. That did work Indeed !!
Much Appreciated !!
Amit Gupta
-Original Message-
From: Mohr Jr, Richard Frank (Rick Mohr) [mailto:rm...@utk.edu]
Sent: Thursday, January 08, 2015 11:39 AM
To: Gupta, Amit
Cc: Mervini, Joseph A; lustre discuss
Subject:
Amit,
Are you by any chance specifying a unique value for --fsname for each of the
MDT/OSTs? If so, keep in mind that fsname needs to be the same when you format
all the MDT/OST devices and that the fsname doesn't have anything to do with
the mount point for the MDT/OSTs.
--Rick
On Jan 8,
In lustre you MUST have a MGS _AND_ MDS node and an associated MDT (file
system) running on the MDS. Typically the MDS and MGS are configured on the
same node.
If you don't have this, lustre won't work.
Joe Mervini
Sandia National Laboratories
High Performance Computing
505.844.6770
Joe, I agree and I have it configured the same way as you described.. Both the
MGS and MDS are on the same node. Adding the output ..
What am I missing here ?. Thanks
Here’s what I had used on the MDS/MGS node to configure the lustre fs
mkfs.lustre --fsname=temp --mgs --mdt --index=0
That is starting to make sense.. Let me summarize so I have got it right
Current
mkfs.lustre --fsname=temp --mgs --mdt --index=0 /dev/mapper/mpathcp1 - On
MDS/MGS
mkfs.lustre --ost --fsname=client1 --reformat --index=0
--mgsnode=10.177.33.10@tcp0 /dev/zvol/LSD1Pool/LusterVola0 - ON OST
New
