SCIENTIFIC-LINUX-DEVEL Archives

December 2007

SCIENTIFIC-LINUX-DEVEL@LISTSERV.FNAL.GOV

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Troy Dawson <[log in to unmask]>
Reply To:
Troy Dawson <[log in to unmask]>
Date:
Thu, 6 Dec 2007 11:01:43 -0600
Content-Type:
text/plain
Parts/Attachments:
text/plain (45 lines)
This is a notice to all those running Scientific Linux 5.0 as a xen host for 
virtual machines.

The kernel security update (kernel-2.6.18-53.1.4.el5) that went out earlier 
this week contained changes dealing with how it deals with the xen host. 
Because of these changes, if you update your kernel and reboot into this new 
kernel without changing xen, you will not be able to host any virtual machines.

Today we have sent out the new xen and libvirt that work with the new kernel.

The problem is that the kernel is *not* automatically updated by the nightly 
yum cron job.  xen and libvirt *are* automatically updated by the nightly yum 
cron job.

What should you do?
Update both as soon as possible.
How?
   yum update
   (check whatever you want to check)
   (shut down all your currently running virtual machines)
   reboot

If xen get's updated before the kernel changes, will my virtual machines die or 
go crazy?
In our tests the answer is no.  Both fully-virtualized and para-virtulized 
machines stayed up and stable after the xen was changed.  But we could not 
reboot them.
Once one went down for whatever reason, it stayed down.  But the others stayed 
up and stable even when several other virtual machines were brought down.

If your xen is changed before the kernel is changed, you cannot start any more 
virtual guests.  But any currently running virtual guests will remain running.

Please note: We were not able to test for long periods of time.  So please 
don't put off rebooting into the new kernel for a month or two.

Thank you for your cooperation.  We apologize for any inconvenience.

Troy Dawson
-- 
__________________________________________________
Troy Dawson  [log in to unmask]  (630)840-6468
Fermilab  ComputingDivision/LCSI/CSI DSS Group
__________________________________________________

ATOM RSS1 RSS2