Another item to talk to first is the idea of performance. So we saw before some of those profiles. So it's one thing to get Nagios to do, to use plugins to be checking things like those buffer cache shits. Here in the performance configuration, this is where groundwork is as part of that profile, giving you what you need to know to build an RRD from an advanced Nagios check. And an RRD is a way of essentially keeping track of historical information, right? And then one would hope making charts from it after a while to keep up with historical performance. Correct. It's an efficient way to store time series data without eating up lots of disk space. Right. So if I'm checking buffer cache hits every five minutes, that's X number of checks a day, number of weeks, that's a lot of rows for the database. Absolutely. You know, a nice thing about RRD is I can use a fixed file size and just update those values but still be able to produce a graph. Let's actually view that as an example. So now I'll come in and I'll ask, well, where am I running that check? Here's that service check. And here I can see I've got a couple of hosts that I'm getting those measurements. I'm going to add those two hosts to the view. And at first glance, let's just do an expanded view. So let's see over the last one day which machine is running different. So you'll see I have a little outage in the lab there. Right. But over the last day there's that value. Now for capacity planning, I can now also take those two graphs and I'm going to combine them now. So now I can see which ones are running hotter than the other. But let's add some other values for capacity planning.
2:00I can come back to this application. And again, because I've used the groundwork profiles, I didn't have to spend a lot of time thinking about what checks should I run, what thresholds they should be, and how do I build the URRGs. I'm going to add these two. So that adds disk on those same hosts. Let's also add memory to that. I'll grab those two hosts. And now we'll look at that same collapsed view, or consolidated view. And now I can get some interesting correlations. All right. So disk space. Okay. One is using 80% of disk. The other is using 50. But if I'm on the hunt for physical servers that I want to virtualize, well here I can see that for the most part these servers are using well under half their CPU and memory. As I recall, in 5.2 you added BERT integration, right? And is that something else or is that what we're seeing here? So these are actually just the RRDs that we're looking at. Now, BERT extends some reporting capabilities. And we have in 5.2 created the ability to put data into the database as well as into the RRDs. And the value there is things like top 10 reports. Right. If I have all of my disk space inside of RRDs and I want to see well who's using the most disk space, that's a tough thing to get out of an RRD. Let's go ahead and pull an example report based on a host group. I'm going to choose my WinXP host group for a second. And let's see what performance data we have available. Now, with some labeling, you can make this a little clearer. But here's the C drive that we're checking. And now we want to just produce a report over time for that particular group of hosts and that particular measurement. So there's sort of a distinction between your more up to the minute or up to the whatever time interval sort of performance graphing and reporting versus the data warehousing kind of reports that you would do.
4:05The data warehousing ones that take sort of a higher level view of all of your infrastructure. Yeah, one is more for, again, analysis. And then the graphing is more for response. Right, right. So it gets that warning that disk is at 81% use. They want to see, well, when did this thing start to run out of space? I just need to see visually a trend. Right. As to now the guy who needs to go justify the new SAN upgrade, he needs to be able to pull a report that says, well, here's where we're using it. And that lends itself more toward the performance reports. So here I've produced a quick report that's summarizing that all of the hosts are between 80 and 81% used. In fact, it's just the one host. And now if I wanted to go further to get a further drill down on that particular host, the BERT reports allow me to feed the links into the data itself. So a drill down now, I'm looking at that specific host. My trend would be there over time, minimum, maximum. And then if I wanted to export this, upgrades in 5.2 to the BERT engine allow us to now give you Word or PowerPoint or Excel even as an expert. Now as another example again, so availability reports, some other extensions in working with customers both for host group availability. So again, with some of those checks that we were looking at before, if we wanted to get availability stats for, say, core routers. You can define, if you have an SLA in place, what constitutes whether or not hosts are up or down. And that's always the trickiest thing with an SLA. Well, if you're up 99.9% of the time, well, is that the whole day or is that minus the scheduled downtime? Right. Because you have to be able to schedule things for downtime. You can't possibly be up, you know, 5.9 that easily. So here out of the box we were able to hone in to see about, as far as the services the core routers were running, we had some warning time that was unscheduled.
6:11And then when I want to drill down further, the hosts themselves were all up. So if the SLA wants to say, hey, the routers can be down for an hour a month, you can actually have that report come through to say, yep, we met the uptime guarantees. Right. And then if I want to drill further into it, I'm able to see some other types of the other outputs broken down for me. So 20% service availability versus host availability. Hosts are all up 100%. Services were up 80%. The last thing is to just speak to the dashboards. So as I have, again, users that need to view a subset of the environment, the idea behind dashboards is that you can collect together specific things that can either help you evaluate current status or you want them to help you troubleshoot. So here's an example of a default dashboard. And you'll see I've got no troubled hosts. I've got some services. I can drill down into any one of them. Mm-hmm. Or if I wanted to get into a breakdown to see, well, who's scheduled to be down and so forth, I have that, as well as the console below. So any of the recent events that have come through can be visible. And is this another thing, like the reports we're talking about, where you can set up a user to be able to just come in and look at this only if they want sort of a self-service view on things? Or is it more intended for the operators to use this? Well, it's more intended where role-based is needed. So we have developers are going to be responding to notifications. I only want the developers to see their web apps and databases. Rather than give them the full status view, I can give them a custom dashboard that can relate different components for them. In contrast, actually for this presentation, I started to build an executive dashboard. And here's an example where I'm collecting together specific items as well as livening it up with a background. So here's a background image that in the dashboard configuration, I'll show you how that works.
8:16But I'm collecting together a high-level roll-up along with some critical host groups and then maybe some services. Right. Where I haven't defined any. So if suddenly I'm logging in because I just got an alert, I want to know, well, what's the status of the critical host groups? Well, if I'm a network guy, I can see on my dashboard all my stuff is green. I don't necessarily pay much time to it. But if I wasn't, if I was a server guy and I now needed to see, well, why is that disk yellow? With a drill down, I can get in there. Now, again, these, again, it's very easy to customize. Sometimes people hear, oh, I'm going to customize open source. It's really pretty easy. So let's, for example, let me go back to my executive dashboard. And I started to open. I can set up roles to assign this dashboard to everyone or just to a specific user or group. So here I am in the dashboard configuration. And I started to have some services be created. So the widget list here is where I can choose elements to add to the dashboard. In the case of this service list, I'm just going to remove this. And I'm going to add that widget back in. So I right clicked on the service list there. And I said I wanted to delete that particular dashboard component. And while it loads, Michael, the other thing I would say is the dashboard builder, we can also embed things that are not groundwork in here. So if you wanted to have, let's say, a part of your dashboard be a page that points to, like, Google Analytics, for example, you can do that too. Oh, right, or maybe your help box inbox or something. Yeah, or Jira or something like that. Right. That's that external URL widget that David was mentioning. So you'll see here when you add that by default, it's actually loading in the groundwork support site. And then for my service list, again, as part of 5.2, we added some additional ease of use type of functionality on the dashboards. One particular thing to note is the ability to use database lookups when I'm configuring.
10:19So now let's configure this for the services that I want to see. If I want to specify a host, let's say it's something on that host we just added. So that CentOS machine is running a service that I want to monitor. And we actually want to see that that particular web check is available. So I click CentOS. I choose the HTTP live service check. And then I choose add service into the widget. And then I can click OK. And notice that that shows up right away with the current state. When I've done my changes to the dashboard, again, here's where I would make it global. Or if I wanted to just set it up to a specific user or a role, I can put permissions on it that say that these users are allowed to read or write. And again, avoid somebody having too much privilege in the system. Being overprivileged burdened. They don't want that. Yeah. And you know that a little knowledge can be dangerous. That's right. So as part of our, again, part of the upgrade to the event console, we also did some work under the hood. Because we had a need for a lot of customers that wanted integrations with service desk applications. Things like Remedy. So we can now do things like take external actions through an API to maybe not just acknowledge all of these problems in Nagios or open all of these log messages. But actually file a communication, say, to a service desk to automatically open tickets or automatically close tickets. Right. So you don't have to switch context to do that. It's just integrated into the tool here. It is. And for Groundwork to maintain some customizations going forward, it was necessary for us to put an API somewhere so that we wouldn't have a series of kind of runoff type of situations. Right. So previously we would implement things with scripts. Now we can write things like to an API where actions taken from the screens can be handled in more of an enterprise fashion.
12:21Right. That makes sense. I think that was a nice overview of Groundwork itself and the new features in 5.2. And I would assume one could just go to your Groundwork website to try it out and see how it works for themselves. Yeah. So there's a couple options. You can sign up on our website for a product tour and you can either take care of it. There's one option to take care of a very brief sort of a quick time tour. Or you can sign up for a demo which is going to be along the lines and it might even be Rich doing it. It will custom tailor it to your particular needs. And then of course we have the Groundwork Monitor Community Edition which you can download as well and play with it yourself. That one's free to use perpetually if I remember right. That's correct. I mean that's the whole classic divide between a community and an enterprise edition or a pro edition. The second one being one that you pay for and have support for and usually have extended features with. Right. Yeah. So in our particular case we have the community edition which is the free one and it's community supported. And then we have the professional and enterprise which are the paid subscription versions. Right. That makes sense. Well great. Well thanks to both of you for taking the time to show us this. It's always exciting to actually see a demo to match up with all the stuff we hear about different stacks and pieces of software. So I appreciate it.