We run a DR test approx 3 times a year. We have never done the full fail over, we just run the DR test in parallel at a remote site.
As far Jesse's question, we haven't explored full failover and going back to the primary site. We probably could with our replication solution but currently it's in the "too hard" basket. We allocate staff to man the roles from user land and our own tech people. In theory the test is a test of our procedures (documents). Most time people flounder with items such as : "how do I clear the screen" (3270 clear!). "What's the session manager jump key". Vendors vary in their response to getting DR software keys. CA are pretty good, although their web site seems to change frequently and between events we find that the place to order keys has moved. One vendor said we should give a few days notice! We pointed out that "we are testing you on this". That usually straightens them up. Our DR is complex, AIX, Wintel and z/OS plus z/VM. The mainframe piece is often quick and then we wait for the toy boxes to be ready for the 3-tier architecture boondoggle. On Fri, Jun 2, 2017 at 8:24 AM, Jesse 1 Robinson <[email protected]> wrote: > Never got traction on two of my questions, which are independent of > technology. > > -- During a failover (test I would presume), who actually performs the DR > procedure whatever it is? Sysprogs, operators, production control folks, or > someone else? Has anyone dared to bring in a non-technical person like a > manager? This question is crucial to business resiliency because, depending > the reason for failover, your top technical folks may be indisposed for an > extended duration. > > -- If you stayed in the DR environment long enough to have > captured/updated live customer data, how did you eventually return to the > production environment? This question is crucial to business resiliency > because at some point down the line, you have to return or, as the poem > goes, settle in for a long winter's nap. > > . > . > J.O.Skip Robinson > Southern California Edison Company > Electric Dragon Team Paddler > SHARE MVS Program Co-Manager > 323-715-0595 Mobile > 626-543-6132 Office ⇐=== NEW > [email protected] > > > -----Original Message----- > From: IBM Mainframe Discussion List [mailto:[email protected]] On > Behalf Of Tony Harminc > Sent: Monday, May 29, 2017 6:10 PM > To: [email protected] > Subject: (External):Re: DR Failover > > On 29 May 2017 at 11:02, Jesse 1 Robinson <[email protected]> wrote: > > > So I'm wondering about other shops' experience with mainframe DR. Has > > it ever been necessary to keep the business running? Who orchestrated > > the procedure? How long did it take? And finally, how did you get > > production back to the primary data center? > > > I have plenty of old war-stories from 1977 (heh - 40 years ago in Feb) > when a fire forced us out of our building for a week. But most of those are > barely relevant to today's world; not just that the technology has changed, > but that we didn't *have* a DR plan, and relied on the good graces of IBM > and (the then monopoly) Bell telecom to get us on the air temporarily. > > It was a lot of fun in its way, everyone learned a lot, and our ops > manager dined well at SHARE and other places where he gave his fire slides > talk for several years after. > > Certainly one now blindingly obvious thing we learned that hadn't really > entered anyone's head at the time is that DR is at least as much a business > and organizational problem as a technology one. For instance, in those days > before mobile phones, and when the closest thing to email ran on the down > mainframe, where do you *go* when your office building is surrounded by > fire trucks and a lot of yellow tape and fast-freezing water, and there is > no one to answer any of the work phones? > > Anyway - the old stories are probably more of a Friday topic than anything > very useful in 2017. > > Tony H. > > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to [email protected] with the message: INFO IBM-MAIN > -- Wayne V. Bickerdike ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
