What is our primary use case?
We use it to monitor about 500 instances, 500 servers, and it keeps up with everything. I monitor Foglight. I wake up in the morning and it's the first thing I look at, because I can depend on it.
We have it on VMs in-house and it performs very well.
How has it helped my organization?
The solution has nothing to do with how the organization functions. It has a lot to do with how I do my job, and how I can help the organization stay on top of things. I need to know, from the ground up, what's going on and when it is happening. The tool allows me to know that. I don't know that the company realizes the value of it. Of course they do, because they're paying for it, but the DBA team, for sure, knows the value of it.
It allows me to do more than I could otherwise. The tool does your job for you in a lot of ways. If I had to collect all that information myself about 500 different instances, I'd need a year to do it. It does that every day. Now, I'm free to fight those fights and still have time to do the upgrades and do the other things, the fun stuff, including building stuff, instead of just troubleshooting all the time. It brings with it a growth factor too, for any DBA that wants to show their value. Just watching a monitored screen is not going to provide much value, in the eyes of the bosses. But when you can do that and do upgrades and other stuff in a week, then you start to show value. It provides you time to do everything else because it does so much.
We also get emails from Foglight every day about long running queries, long running jobs, and broken jobs. Again, it's really doing my job for me. I just have to respond to it. It tells me what I need to do and I do it.
We can also drill down and do root cause analysis for most things. That's a huge benefit of the tool. We mostly have Windows boxes. Between the drill-down for root cause and Event Viewer, you can decipher what the root cause of anything was, or just prove what it was. The same information comes out of Foglight that would come out of Event Viewer. It saves us tons of time. I couldn't do all of my job in one week if I didn't have the tool. The company would probably have to hire another person if we didn't have it.
What is most valuable?
We created a dashboard called "Morning Coffee," and when I'm having my coffee in the morning, everything that has happened, good or bad, shows up on that dashboard. That's my favorite, because that's where I make my money. That's where I show my value to the company, because when things start to tilt in the wrong direction, I know it. To me, that's huge. You could talk about the emails that come out that tell you the server is down—any application can do that. But to collectively have information for all 500 instances, at my fingertips, is huge.
That dashboard is custom made. The gentleman I work with, Brant, actually created the dashboard. He has a section for failed jobs. He has a section for failed backups. He has a section for servers down, which hopefully is never populated. Everything you'd want to know about anything that happened while you were sleeping is there, and it's actually there for however long is necessary. It could be 24 hours. It could be two weeks.
Also, it never seems to fail on the alerts. The alerts are solid.
Foglight's Change Tracking capabilities are another huge feature. It is wonderful to be able to do that. People don't realize the amount of information that Foglight gathers from a given server or instance. It gets down to the version of SQL, the disk array, everything that's there. Any change that's made, any upgrade to SQL, shows up on the dashboard almost immediately. You don't know how much time you save just having a tool in your back pocket that does that for you.
The solution also provides real-time activity screens. You can drill down into real-time for
- disk space
Anything you want to see is there, and there's a drill-down for each server that shows you that information. It's a separate page that comes up, and shows you, like a dashboard does, everything at once. Then you can drill down further into anything there that might show a problem or a problem that is about to happen. That drill-down feature and the ability to see everything that's going on, on the server, is a really nice feature. It's great because you want a screen to show you stuff before the end-user screams, and this feature allows you to know that information. With it, I know when things are going awry before the end-user does, and you can't ask for more out of a tool.
It also enables us to monitor multiple database platforms. We monitor, in-house, something like eight versions of SQL, most of which are on different OSs, different servers, different hardware. We're not doing Oracle yet.
The Performance Investigator feature in this tool is really good. We only use it for production, but it drills down to the narrowest bit. If you were to log in to my system, I could tell you, the next day, when you logged in. If the SA account logged in three months ago, I can tell you that. The SQL PI is really a huge feature in the tool.
What needs improvement?
One thing that I would like, and it's probably something that I could set up internally, is something other than a dashboard which I have to look at to know that a server is down. I'd like bells and whistles to go off. While the tool allows you to prioritize those, based on the severity of the server—if it's high-level production or low-level production—I'd like to know, by having something tell me, if I'm not in front of the screen, that I have a server down.
If I look at the dashboard I know there's a server down. But if I'm not looking at it, if I'm looking at some other problem, I want to know about it. You can do that, you can use SMSs and alerts to your phone, and I could set it up to handle that, but it would be nice if, out-of-the-box, Foglight did that.
For how long have I used the solution?
I have been using Quest Foglight for Databases solidly for two years. Prior to that, I used it off and on for another two to three years.
What do I think about the stability of the solution?
The stability has been awesome. In the last six months we've had one slowdown, and it was easily resolved. We rebooted the environment and it went away. That's one hitch in two years.
What do I think about the scalability of the solution?
It does a much better job, when it comes to scalability, than Grid Control. We've got 500 instances, just SQL, on there. That's huge. I can see that it would handle another 500.
How are customer service and technical support?
Their technical support is great. They are super-fabulous. You open up a ticket and someone always gets back to you either by email or on a phone call. They're really good. I'm an Oracle guy, and I used Oracle support forever, and it's tremendous what Quest does as far as support goes. It's their standard support and it's wonderful.
How was the initial setup?
I wasn't here for the brand-new implementation. They've been using it for longer than I've been here, but I've been adding to the environment as we go along.
When you introduce a new target, a new server into Foglight, that is really straightforward. They make it so simple to do and it does all the work. You say, "This is what I want. This is the name," and it goes after it, and it installs agents everywhere they need to be on the OS to launch the database. It's a two- or three-minute process, if that. That part is wonderful.
For maintenance, for our environment, we need two DBAs: one full-time, and one helper. That's how we have it now. Brant and I handle the environment. He's the lead, and I'm his backup, but I'm there every step of the way. The two of us use it 100 percent every day. We have six or seven users of the solution and, if you include management, there are probably 12, as we have that many accounts in the tool. All of the users are DBAs.
What was our ROI?
If you don't have this tool, you need at least another body. If you think of the going rate for a DBA, and at least one or even two of them, annually, that is ROI.
In addition, you're not going to get the work done, work that the tool does for you, before you even wake up in the morning. It really is immeasurable. If you've never had Foglight, you don't know. But if you used to work without it, once you've used it, I'm not sure you'd want to do your job without it anymore.
Aside from the alerts, the emails you get in the morning, and the alarms that go off, it's the collection of data that is valuable. You can go back to any time you want, pull a report out and hand it to somebody and say, "This is what this CPU looked like for the last year and half, and we need help with it." If I had to go collect that information, it would be totally impossible. I don't know how many people you'd need to do that.
Which other solutions did I evaluate?
I'm an Oracle guy and I've used Oracle's Grid Control, which is similar to Foglight. Foglight offers so much more. I was an instructor and I actually taught Grid Control. While Grid Control is good, I'm not sure it's as stable or as powerful, but it's good. It does the same type of thing. It handles a server with any of the databases on there.
One thing that Grid Control does, and I'm not sure if Foglight does this—and it would be a nice-to-have—is that it enables me to pick out two or three servers in my environment and do a comparison among them.
If Foglight had that, that would be really nice for a multitude of reasons, one being licensing. Thinking it through, there are a whole bunch of applications for that kind of capability. For example, if you're planning an upgrade across the board, what are you upgrading, and why? If you could pull that information out of Foglight easily, that would be great. I can create a report and get the same information, but my point is that, in Grid Control, there was a standard page that allowed me to do a comparison within the application.
What other advice do I have?
The biggest thing I've learned from using it is the reduction in effort that is required to do my job. Don't tell my boss that.
My advice is "buy it." You won't know until you use it. I've been a DBA for 22 years and it really is an awesome tool.
We use Foglight to display the most intensive database queries, but it's on a per-server, per-instance basis. We haven't created a dashboard for that, although we probably should. I can drill down into a server and I can tell you, from top to bottom, which queries are the most expensive. It could help us to improve query efficiency but we don't use it that way. We have vendor-supported applications and they're responsible for that. So that's not our focal point.
Overall, it really is a good tool. I think it's the best on the market.
Which deployment model are you using for this solution?