What is our primary use case?
We monitor quite a few database servers. The actual number jumps up and down on a regular basis, but on average we're doing 120 servers at a time. It gives us one pane of glass to be able to see which ones are having actual issues and which ones are just going along.
How has it helped my organization?
When we do have issues, for example, that our financial software starts having slowness, we can use the Performance Investigator module and dig into where it's actually slowing down. It allows us to do the troubleshooting and resolution at least 10 times faster and get the users back to work. I and one other senior DBA on the team have built queries that we can dig in with, but going through all the results is huge and time-consuming. This solution helps us narrow in on the problem a lot faster.
Also, our AppDev team used to love to develop on production servers. By being able to show them the metrics of how they were actually affecting the performance, we were able to get them to move to a development server and not do any development work while they're on the production server. In the four-and-a-half years that we've been using it, that change has probably saved us four years' worth of time.
What is most valuable?
The most valuable feature is that it's one pane of glass and enables us to see everything at once.
The other senior DBA spends a good part of his day in it and he's focusing on indexes right now. The PI module allows him to identify which new indexes or modified indexes are going to make the biggest impact.
What needs improvement?
I had never used Foglight before I got to this company, because I didn't have the time. I had other responsibilities besides just DBA work so I couldn't focus on what Foglight could give me. Having said that, there's still a lot of "noise." I get a lot of alerts that, while important, are not critical. Then I have to dig in and figure out how to turn alerts off, but not the logging. I want to be able to go back, once we get the other big issues out of the way, and start fine-tuning some of those other areas, but I don't necessarily want to receive an email for all of them. Over this past weekend I had 400 emails from Foglight. That's a lot. And at least 395 of them were white noise.
They need to make an interface where it's easier to turn the alerts off but not turn the alarm off. The other senior DBA on staff got frustrated with the alerts, so he just went and turned the alarm completely off. I said to him that while it won't alert us anymore, we'll also lose visibility into that aspect. It's something that we do want to be able to see at some point, just not right now.
For how long have I used the solution?
I have been Quest Foglight for Databases for about four-and-a-half years.
What do I think about the stability of the solution?
I like the stability. The only outages we've had were related to updates. We had a one-day outage when we upgraded and that was due to issues that were not documented in the update process. Otherwise, the product itself hasn't crashed and the virtual machines haven't crashed.
We notified the vendor of the undocumented issues and they were really quick to get on the phone and tell us, "Okay, this is a step we didn't put in the documentation, but we need you to do the following."
What do I think about the scalability of the solution?
It's definitely scalable.
Now that they've moved to the SQL Server side for a SQL PI, the Performance Investigator, it's a little limiting. We've had to increase the drive sizes to capture that Performance Investigator information, but I prefer that it's in a SQL database because it's easier to report that way. Also, previously they were using a product called Infobright which stored everything on the C drive. There were a couple of times where that drive would fill up and once the C drive fills up, you can't do anything with it. I was glad that they moved to SQL. We put it on its own instance so it has its own set of drives and, more importantly, it's not on the C drive. And if we need to, we can expand it.
We'd like to increase usage of the solution if we can. I'm trying to get the application development team to use it more extensively. We also have a new warehouse that opened up and I'm trying to get the person who's supporting the application there to use it. I expect that our usage will expand.
How are customer service and technical support?
The technical support is really great. I've never had an issue that they weren't quick to jump on and get resolved rather quickly.
The sales support is very good. Once they see a request come in, they'll help to escalate it, if necessary. Overall, I've had a lot of really good experiences with their tech support. I'm very pleased with that.
How was the initial setup?
I was the only one involved in the initial setup in our company. There was a little complexity to it, but overall it was very straightforward. We didn't have any real issues getting it set up and running.
You've got to let it run for a while before you determine what is white noise and what are actionable items. Then you have to go back in and say, "This is not something to alert on, but it is something that I still want to log." Sometimes that white noise does come in handy when you're looking at troubleshooting a long-running issue.
From start to finish, the deployment took a week.
First off, I had to get all the servers built and we did virtuals. But I had to get a tie-in with our server team to get those set up and running. The requirements, themselves, were pretty straightforward. I could present to the server team exactly what we needed and how we needed to set it up. Getting the basic infrastructure in place was what took the most time. Once we actually started the install of Foglight, it was pretty simple.
What was our ROI?
I haven't been able to pin down an exact ROI, but I can easily say that it has helped with expenses that would be related to certain issues.
I can give you one really good example. We've got 32 stores that are scattered all around the country. Foglight was able to identify that none of these were being backed up. We got an alert: "Hey, we don't have a backup for these servers." That got us to start backing up those servers. If one of those servers crashed the process was to get a new server put onsite. They would have to start from scratch, install SQL, create a blank database, and then have to spend 24 to 48 hours getting it caught up with information that the system already had, back here. In the process, they lost two days of sales but they also lost the data that was on that server.
We got a call one day that one of the servers crashed. We had a server sitting in the shop and rushed it out to them in an hour. They said, "Okay, we did a restore of the database," and within 20 minutes they were ready to open the shop. They called the store manager and said, " Okay, you're up and running, ready to open the doors again." And she said, "I just sent everybody home. I thought we were going to be out for two days."
We were able to resolve the issue because we were aware of it. That's what I like about Foglight. It does help us to be aware of potential issues and even get ahead of them.
What's my experience with pricing, setup cost, and licensing?
The price is worth it, if you have the time to go through the information.
I have worked with the sales staff at Quest by talking to other potential customers, and have said, "If you don't have the time to focus on the issues that it can present to you, if you've got to split your time between database administration and system administration or helpdesk, then maybe Foglight is too much for you."
There are other modules that you can add in for additional cost. For example, you can do network monitoring tools and I believe there's a physical Windows Server monitoring tool. We don't use those because our server team and network teams both have tools that they like better.
Which other solutions did I evaluate?
In other jobs, I've used other products. I've used the other product from Quest called Spotlight. I've used Idera and Redgate monitoring tools. They're great if you only have the time to look at the general performance, whereas Foglight gives you enough detail to actually resolve a SQL-related issue.
Foglight is a really good solution for database monitoring. With that being said, it gives you the opportunity to get so much information that it's overload, if you don't have the time to dig into it.
What other advice do I have?
If you've got the time for it, the time to focus on databases in general, then Foglight is definitely worth the expense because of the information that it can provide for you.
The biggest lesson I have learned from using this solution is that it's worth it. It enables you to pin down troubleshooting within 30 minutes to an hour, whereas before, you'd be pouring over reports or data from queries for days. That's huge. The CIO has told me that since we've started using Foglight, we've actually gotten ahead of some of these issues and we're actually being proactive instead of reactive.
We're in it all day, every day. I and at least two other DBAs are in it regularly, as well as some AppDev team members that we're trying to get to use it. We've got other database wannabes that are using it and our systems admins use it as well. Overall, there are 10 to 15 users. In the IT department, it is used pretty extensively.
There aren't a lot of tools that I've tried to integrate it with. I'm in the process, when I have the time, of integrating it with ServiceNow.
Which deployment model are you using for this solution?
Which version of this solution are you currently using?