It depends on your approach for DR.. is a physical backup the only method? Typically, it's a hybrid .. use physical backups to restore as much as you can (hopefully enough to run linux and be able to start backup clients to restore) - and then use backup/restore clients to go to logical backups for latest 'clean' data.
Scott Rohling On Thu, Aug 30, 2012 at 9:52 AM, Aria Bamdad <a...@bsc.gwu.edu> wrote: > True, no solution is as safe as shutdown/snapshot/restart in terms of > ensuring consistency but I think it would be nice if there was some > mechanism within Linux that would alert the kernel of this and would allow > for a relatively consistent snapshot of the virtual machine, assuming > perhaps we stop application servers databases on the host prior to > snapshot. > Then, we can at least minimize downtime for the server and reduce the > resources necessary to bring the host back up. I am imagining trying to > shutdown/restart dozens or hundreds of virtual machines to take a snapshot > for DR would consume a fair amount of resources and slows down the process > a > lot. The bigger problem in my mind is that if you set this up to run > unattended, then you have a much higher risk where a server has a problem > upon restart. > > I wonder if anyone would like to share their experiences with regards to > automated shutdown/restart of linux guests for DR snapshot purposes. > > Aria > > -----Original Message----- > From: Linux on 390 Port [mailto:LINUX-390@VM.MARIST.EDU] On Behalf Of Rob > van der Heij > Sent: Thursday, August 30, 2012 11:43 AM > To: LINUX-390@VM.MARIST.EDU > Subject: Re: LGR guest quiesce > > On 30 August 2012 17:11, Aria Bamdad <a...@bsc.gwu.edu> wrote: > > > Just curious, could this functionality be used somehow to allow for a > > **consistent** disk snapshot using flashcopy for a live Linux guest? > What > > I > > mean is that currently, we have to shutdown the guest, do the flashcopy > and > > then restart the guest. Is there a way we can tell the Linux guest to > > freeze, take a snapshot of the disks and then resume? > > > > No. It's perfectly possible that the Linux guest is moved from one z/VM > LPAR to the other while it still has data in memory that's supposed to go > out to disk (and it will eventually do that from its new location). > > Rob > > ---------------------------------------------------------------------- > For LINUX-390 subscribe / signoff / archive access instructions, > send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or > visit > http://www.marist.edu/htbin/wlvindex?LINUX-390 > ---------------------------------------------------------------------- > For more information on Linux on System z, visit > http://wiki.linuxvm.org/ > > ---------------------------------------------------------------------- > For LINUX-390 subscribe / signoff / archive access instructions, > send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or > visit > http://www.marist.edu/htbin/wlvindex?LINUX-390 > ---------------------------------------------------------------------- > For more information on Linux on System z, visit > http://wiki.linuxvm.org/ > ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 ---------------------------------------------------------------------- For more information on Linux on System z, visit http://wiki.linuxvm.org/
