Transcript
you an actual live role swap with RobotHA. Bob and I work with a lot of customers and we find that they're just not simply doing role swap testing and we're here to show you how easy that can be. So Bob, over to you to talk about your background and why you're an expert in this area. Hey, Tom, thanks. Yeah. My name is Bob Butcher. I'm a solutions engineer here at Fortra. Prior to Fortra, I worked for a managed service provider and I was also a customer and used the RobotHA product. I've been on the platform, the IBM iPlatform for 30 plus years. Awesome, Bob. And myself, I'm our EVP of technical solutions at Fortra. I've been training customers on journaling, backup and recovery. I'm an IBM Power Champion. It's been my role in the marketplace to help customers understand automation with backups, automation with IBMI. I'm very excited to be with you today to talk about role swaps. Before we dive too far into that, there's a couple of things that you need to understand. One is, HA, DR, are they the same thing? They really aren't. High availability means where you can instantaneously switch over or in a reasonable, very quick amount of time, you can switch over. So high availability might mean something, a solution just in your data center. Whereas we talk about DR, we're talking about a data center or recovering your system to a remote site somewhere and that also can play into a product like RobotHA or it could even, for some of you, mean from tape and restoring your system from a major disaster and maybe tape is good enough. So that gets into our next topic, which is RPO and RTO. Now, RPO is just simply how far back do I have to go to recover my data or redo my data after a disaster happens? So if a disaster happened at four in the afternoon, do I have to go back to my backups or do I go back to 359 and 59 seconds because I'm using an HA solution that is replicating my data near real time? And RTO is recovery time objective. How quickly do I need my business up and running? Those are very important points as we think about role swaps and bringing systems up for DR, HA reasons. Now with RobotHA, we have a test while active role swap, which basically is that your production server is up and running while you're testing your backup server. Could I use the data that's over there? We're not going to spend much time on that today. We're really going to do a full role swap with you. Bob's going to show you a live demonstration here of actually doing a full role swap with RobotHA. So Bob, we're going to take it over to you to do this live demonstration and we'll do some commentary along the way as this happens. So if you want to take it away from here. Yep, I'm going to take over your screen, Tom, and let me jump over to here. So while he's doing that, I'll remind you that RobotHA has the ability to do a test while active role swap or full role swap. What we're showing you today is going to be a full role swap. Bob's actually signed into his console devices. We do require you to run this from a console. So here we are on a system called Fred and a system called Barney. Fred is the production side. There's many different names for it. And the Barney side is the backup DR server, HA server. It's called all kinds of things from different customers or target server is another term I hear. Go ahead, Bob. Yeah. So the other thing too, Tom, what we've done is we've kind of given it colors, right? We're going to do a swap IP. So basically what we're going to do is we're going to take the Fred system, move it over to the right side where Barney is and vice versa. But we've given you colors in the upper right-hand corner. You can see Fred is my prod, my star prod in the upper right-hand corner. Barney is my backup system on the right-hand side. And I logged into my RBO menu, option 25 is RobotHA. I'm going to go down to seven on the production side and I'm going to do a role swap. Now when I press that, I'm going to do an option 10. Option 10, what Tom was talking about, set the test flag. If I wanted to do a virtual role swap, I could put yes there, but we're going to do a live role swap and I want to make sure we are going to do a star yes for are you sure my current role is star prod of the box. And I'm going to do this. This kind of matches with the IBM, the F16 when you're doing your IPLs. So we're going to start doing this now in the prod system. So we don't want people to just inadvertently run a role swap, right, Bob? So we want to make sure that you know what you're doing there. That's the purpose behind that. So what are some things that you need to think about when you're getting ready for role swaps, Bob? I know you talk about sync attributes. What are those? Yeah. So a lot of times, Tom, when we do demos, we focus our attention on libraries and the IFS and things like that. But some of the critical things that should be checked are your user profiles. How often are your user profiles sent over to the other box? Your configuration. If you've got a lot of things with you're dealing with the web, system values, job schedule, network attributes, those things are just as important to make sure that when that other box comes up on the opposite side, that you have the same look and feel that you had before you did the role swap. So those are important. So right now you've started the role swap on the production side, the source side, your business side, right? Very good. Yep. Yep. Yeah. So, Tom, you were talking a little bit about the RTO. What is RTO? Why does that matter? Again, yeah, RTO again is recovery time objective. So in other words, how long before your business is going to really suffer that you haven't brought your system back? So when we think about backup to tape, recovery time objective, if you're really good at recovering from tape, you're probably looking at a minimum of eight hours. But for a lot of customers, it's 24 hours, it's 48 hours before they actually can recover their system. So what we need to do is look at what is the business cost of being down for an hour, for example? What's that really cost you? I know, Bob, you had some examples as a customer of what you found it cost you for an hour of downtime. What was that? Yeah. So the first time we actually did this and we recovered our system from tape, we were down for a couple of days and we figured out it was well over $200,000 an hour. So that's when you go to management, it kind of justifies the cost of maybe an HA solution for you. Right. And you guys were a mid-sized business, you're running what, five terabytes or something like that, if I remember right? Yeah, we had seven terabytes. So seven terabytes on a production server. So nice sized system, really. Yep. Yep. Okay. So now what you're going to do is he's going to start the role swap over on the backup server too and get that going. Bob, you know, one of the things we didn't mention earlier was how did you know you were ready to do a role swap? And that the system was, it's a good time to do that full role swap because there are some things that you need to do ahead of time as a customer, right? Yeah. So one of the things that I took advantage of, I've already run it now and I can't show the report, but there was an audit history report that kind of talked about what passed, what failed. If you were doing a role swap right now, what's going to work, what's not going to work. And I would religiously look at that every day because I'm like, I want to make sure that if, you know, God forbid we had a problem, we could do a role swap. And some of the things that it showed is everything was in sync, the configuration objects or the user profiles, the authorities, things like that were okay. And it would just show you that you were good to go. If you wanted to do a role swap at that point in time, that you were ready to do that. So why don't you go ahead and hit enter, I think, or the command key on the back of the box. And we're going to do the same thing, Tom, on this right-hand side. Are you sure we're going to do a star yes, set test flag. We're not doing the virtual swap, which is star no. I'm going to go ahead and I'm going to press enter over here. Again, I'm going to press F16, that's that confirmation. And so now we've got both boxes. And what's going to happen, Tom, behind the scenes is we're actually the swap type of swap IP. In this case, we're on the same subnet. So we're just going to swap Fred and Barney in the different spots. And you can see that which requires an IPL. So we're going to see go down and come back up. So the system names are actually going to swap in the IP addresses and all those things in the background. That's not normal, by the way. Most of our customers that we work with, their DR box is out of their subnet, which is a good thing. And so we're not even swapping IP addresses with those. So there's no IPL required. But if you're going to change your system name and your IP addresses and stuff, we need to swap it. Now, looks like we're done over on the left side, on the production side, and it's actually switched it over. It's actually going to do a little IPL over there. And one of the things, too, we talked about being prepared and ready for a role swap. And our auto report's really good for that. But let's just say you're a customer out there that you have a really limited staff, or you inherited an IBMI server, or you might call it an AS-400, and you don't know what to do with HADR. One of the things that's nice with RobotHA is that we have a program called RobotHA Care. And depending upon how much touch you want from us, we will help you out on a monthly basis, a weekly basis, or even a daily basis, and we'll even help you with changing rules within RobotHA if you're in the full RobotHA Care, where we basically are taking care of and monitoring your system and doing remediation for you. Now, that might be for the larger customer, but the smaller customer, quite honestly, if we check your system once a month, the software's reliable enough that that would be a good way to go. We can even set up things like email notifications and items like that to help. But we'll use some of those same technologies and give you that care. Now, the most important part of this, we'll coach you through doing these role swaps, a full role swap while active first, and then a full role swap at the end of the day, which is ultimately what's important. Awesome. And actually, as a customer, Tom, I took advantage of that. I had the support staff kind of walk me through that stuff until I was comfortable with that. One of the things that I used, I did my backups in the IBM license program, the BRMS, but you could do it in RobotSave too. I use something called a recovery report. I would run that on a daily basis too. After my daily backups were done in BRMS, I would run my BRMS maintenance, and I would generate a recovery report. What that showed me is what volumes would I need if I had to recover my system. Now, why that's important to me is because when people think about a role swap, they think, well, I'm going to get my system running over on the target side. I'm going to get it up and running on the target side. If you actually went through a major disaster, let's say your data center was on fire, you flooded, like in our case here, we had some issues with flooding, you need to know how to recover that production box, right? You might have to call IBM, they might have to wheel in a new system, you might have to get new hardware. That BRMS recovery report, or if you're using RobotSave, will help you restore that system. A lot of people don't think about that, so that's why that's important, I think, that recovery report. Now, this is really exciting. What's happening in the background, Bob, with these two systems? Yeah, so basically, the systems have IPL, and the reason why they have to IPL is because they're actually changing the name of the system, right? Remember, Fred was on the left side, Barney was on the right side. In order to change the name of the systems, we have to do an IPL, right? So behind the scenes, it's taking all of that information, and it's changing both systems. So on the left side, you can see that now Barney is up, okay? The system that we started on the right side is now Barney on the left side. So we are in the process of doing an IPL now, Tom, on the right side, and that system should come up as Fred. So it should be up here shortly. Yeah. So, you know, role swap testing is really important. You have to be able to restore these partitions and be able to restore them in a reasonable amount of time. And again, if you're not comfortable with that, that's where our RobotHA care comes into play. But we also think it's important, even though you have technologies like RobotHA or hardware replication, whatever, you still need to test recovering your tapes or recovering from tape type backups, which might be virtual tape libraries or actual tapes, either or. You know, I say, the thing I like about actual tapes in today's world is they're actually free from ransomware. If you think about it, right? So they're disconnected backups, right? They're not connected to your network. So if you did have ransomware, and it's an IBMI partition that had a backup tape operation, you could restore your system from that backup tape and would be safe from a ransomware attack. But you need to test it, right? Can I restore that system? So now we see here on the right side, we're IPL on that partition, and that should come up as Fred here shortly. You know, going back to testing, how often did you test your role swaps as a customer? So what we did, Tom, is we did two tests a year. We did it from an audit standpoint, you know, we showed the auditors that we would test this plan. And you can see on the right hand side, now we've got Fred, right? So now we flip the spots. You can see that role swap probably took seven, eight minutes total. But when we tested, it wasn't just me, we had the network folks involved, we had our application people involved, we had our operators involved, you know, maybe we need to hold some jobs at the end of the day, we would test all the processes to make sure the applications came up, they were running okay. You know, normally, like we did a swap IP, but in a probably in our environment, we did a like a non swap IP, where now you got to get your network people involved to maybe change some DNS names and things like sure. So it was important to test that annually, and at least annually, again, we did it twice a year. And the reason why, because we went through an actual disaster, when you have an actual disaster, that's the last time you want to worry about testing this stuff. Because believe me, I was pulled in 10 different directions that first hour that the disaster was declared. And this is the last thing that you want to you want to do, you want to know that you've got a process in place a runbook, whatever, to make that go smoothly. Yeah, why don't we go and have you sign on to each of these systems. So we can see the robot ha menu in the background. And, you know, just to further that point of testing, you should also think about IBM is a database server, and that it's serving up applications, web applications, green screens, all those things. But you also have other business applications on other servers, non IBMI, Windows, Linux, Unix, that might be dependent on that data. And making sure that you test some of those connections, or all of those connections, I should say, say is important to as part of your actual role swap. And I think Bob is showing you now that we're now Barney's the backup server, Fred's the production server, we've switched sides here. What's your final point about a capability of robot ha before we wrap this up? The role swaps all done, by the way, people, I know you were very excited about this. Yeah, so so a couple things about robot ha. First of all, test the role swap, make sure it works, right, we just kind of went through and you can see how fast it is. Another point I'll make Tom is migrations. Okay, when I had the robot ha product, I think that's back when we had power eights. And IBM's coming out with the power nine technology and everything else, I actually migrated my systems to the other hardware. So I didn't have to worry about all this downtime and restoring from tape. So so that was important as well. Not only did we want to make sure that the role swaps work, but we wanted to, there were other uses for it with the migrations as well. Yeah, I think about finally, yeah, finally, just the robot ha care basic robot ha care robot ha care plus our programs that we have here to help you out. And then don't forget fortress here to help you out with many different product offerings in the IBMI world. We're experts in high availability and backup and recovery like we talked about. But don't forget about cybersecurity of capacity planning, scheduling, business intelligence, document management, we do all those things on the IBMI server, I want to thank tech channel for allowing us to bring forward how easy it is and how controlled a full role swap can be with robot ha. Thank you, Bob. Thank you for an excellent demonstration. All right. Thanks, guys. Take care.
