On Sunday, 10/21/2007 at 11:45 EDT, Patrick Spinler <[EMAIL PROTECTED]> wrote:
> > Some folks have seen Romney White (one of my colleagues) demonstrate at > > SHARE some z/VM technology that relocates a running guest from one z/VM > > partition to another without an outage. Even though it was a prototype, > > it demonstrated that z/VM is capable of fulfilling the "zero down time" > > for your guests and we are busily working to bring that to a theater near > > you. > > We're also aware of this, and we're very much looking forward to it. > Can't really trumpet this as an advantage, yet, though, since it's not > yet released. Fair enough. > Which does bring up the point, though, that our shop, with more than > one lpar on more than one cec in more than one datacenter appears to > be in a distinct minority. In our local z/vm lug, at least, it seems > like very few other people are doing anything comparable to us. > > So, for the majority of zLinux users, my point would still stand -- > upgrading or patching z/VM takes down your entire linux environment. > There's little or no granularity or rolling upgrade options for most > customers for this. Running a single z/VM LPAR is a customer choice. If you want to improve the availability of guests on z/VM, you need to operate a hypervisor cluster or its moral equivalent. Of course, that's just part of the HA story. To reach the highest levels of availability you need application-specific clustering solutions (e.g. Domino db replication or DB2 HADR), along with more generic solutions such as sprayers and IBM Tivoli Systems Automation (TSA). Which you choose depends on the app, of course. You could easily imagine that when TSA detects ServerA is down on System1, it starts ServerA on System2. That is so much simpler when you're using two z/VM LPARs with shared dasd, a shared directory, and CSE. So I would say that if upgrading or patching z/VM takes down your entire Linux environment, you've got more work to do, or you've reached the level of availability you can afford. Eggs. Baskets. I hope that in the future people will run [at least] two z/VM LPARs as a matter of course, just as they do today with z/OS. Alan Altmark z/VM Development IBM Endicott ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
