If one assumes that the rebuild will incorporate the same identity as the 
original host (same hostname, IP address, etc.), then it should just be a 
matter of restoring the OS, re-installing the Lustre packages, configuring LNet 
(e.g. /etc/modprobe.d/lustre.conf) and remounting. If you've got an HA setup 
(e.g. Pacemaker + Corosync), then you'll need to restore that as well. Or 
rather, keep a backup copy of the config so that you can restore it :). There 
is no need to perform any "rebuild" of Lustre itself; just repair/restore the 
OS.

Other than LNet, all the Lustre configuration information is held on the 
storage targets (MGT, MDT), so you can rebuild the root disks without affecting 
the Lustre config on the MGT and MDT.

So, in summary: rebuild the root disks (maybe use a provisioning system like 
kickstart for repeatability), restore the network config, restore LNet config, 
maybe restore the HA software, restore the identity management (e.g. LDAP, 
passwd, group) then mount the storage as before.


Malcolm Cowe
High Performance Data Division
Intel Corporation | www.intel.com

-----Original Message-----
From: lustre-discuss [mailto:[email protected]] On Behalf 
Of Jon Tegner
Sent: Friday, March 11, 2016 4:48 PM
To: [email protected]
Subject: [lustre-discuss] Rebuild server

Hi,

yesterday I had an incident where the system disk of one of my servers 
(MDT/MGS) went down, but the raid could be rebuilt and the system went 
up again.

However, in the event of a complete failure of the system disk (assuming 
all relevant "lustre disks" are still intact) is there a clear procedure 
to follow in order to rebuild the file system once the OS has been 
reinstalled on new disk?

Thanks,

/jon
_______________________________________________
lustre-discuss mailing list
[email protected]
http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
_______________________________________________
lustre-discuss mailing list
[email protected]
http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org

Reply via email to