You rely on FrameFlow to monitor your critical IT systems but what happens if the system hosting your FrameFlow configuration goes down? FrameFlow’s failover options are designed to handle that case elegantly and easily.
In our 2016.5 release we implemented an important performance enhancement for SNMP monitoring. Let’s look at the details.
One feature in FrameFlow that often doesn’t get noticed right away is the Views section for devices. Find out more in this post.
Have you looked at FrameFlow Multi-Site Monitor and wondered how it works? To learn more about it, watch our two minute presentation. It explains how our unique multi-site technology can help you to distribute your monitoring tasks and scale up your configuration to 1000’s of systems.
CC Image courtesy of jakeliefer (Flickr: Old Keys) [CC-BY-2.0 via Wikimedia Commons]
Techniques for Fast and Efficient Server Monitoring
FrameFlow is highly customizable and offers a wide variety of monitoring options. We’ve designed it that way on purpose to give you the flexibility to build out your monitoring configuration in a way that best suits your work environment, practices and policies.
There’s one area where first time users often have questions and that’s when it comes to authentication. FrameFlow is an agentless monitoring system, which means we don’t install anything on the systems being monitored. Instead we use standard protocols to reach out to the monitored machines and collect the data we need. To do that we need to authenticate to the machines being monitored. This is a good thing because it means that FrameFlow plays by the rules when it comes to your network security polices rather than wildly opening ports to reveal monitoring data like some agent-based systems do.
The vulnerability allows a user with restricted permissions to escalate to domain administrator privileges and Microsoft reports that attacks have been seen in the wild. Microsoft has officially stated that "the only way a domain compromise can be remediated with a high level of certainty is a complete rebuild of the domain," so it's critical that all Windows admins apply this patch immediately.
Microsoft recommends first patching domain controllers running Windows Server 2008 R2 or earlier. Next, patch your Windows Server 2012 and later domain controllers. Finally, patch all of your regular systems to ensure complete safety.
Use the following step-by-step instructions showing how to configure FrameFlow Server Monitor to make sure your systems have been patched:
Anyone running web sites on IIS with SSL support should have already rolled out patches for CVE-2014-3566, the vulnerability in SSL v3 which has been labeled POODLE.
What is the POODLE Bug?
FrameFlow Server Monitor Can Verify POODLE Patch Installation
Did you know that you can use FrameFlow Server Monitor to verify that the patch has been installed on all your systems?
CC Image courtesy of Buster Benson on Flickr
Databases like SQL Server, MySQL and Oracle are the workhorses of IT operations. They are the reliable back-end for applications, services, web sites and many other kinds of systems so it’s no wonder they have to be up, running and available 24 x 7. But despite being workhorses they are still fragile. All it takes is an unexpected traffic spike or even worse, a runaway application, and suddenly response times start to lag, systems start to fail and end users start to complain.
Effective database monitoring means looking beyond system health metrics.
The key to effective database monitoring is to make sure that you are checking the right things and to look beyond simple system health metrics. Don’t get me wrong, system health metrics are great and a necessary part of core monitoring for any system, but even if CPU usage is low, drives have lots of space and memory is plentiful, it doesn’t mean that your databases and the systems that depend on them are healthy.
Two simple techniques can give you deeper visibility and better warning about potential issues with your database server:
First, create sample queries and run them on a regular basis. For effective database monitoring we recommend running actual queries against the database. The main thing you want to check is how long it took the query to complete. Ideally you’ll have a graph of that value over time so you can detect patterns and make an estimate of what normal conditions are. It’s usually a good idea to run separate queries against multiple tables and have different thresholds as some may exhibit different behavior under normal conditions. Set your monitoring software to alert you when the thresholds have been exceeded. Ideally your monitoring configuration will be set to only warn you after several failures in a row. After all, you don’t want to be alerted about momentary spikes, but you do want to be alerted about sustained abnormal conditions.
Second, run health check stored procedures. If your vendor (or internal development team) has not yet implemented health check procedures then it’s time to pressure them now. A health check is a stored procedure that runs internal tests reproducing common database operations that higher level components use. The health check should return a success/failure indicator and text strings indicating what passed and failed, if anything. Run the health check stored procedures on a regular basis and set your monitoring tool to warn you based on what they return.
Get a clear view of normal operations and advanced notice about potential issues.
By implementing the above techniques you’ll get a clear view of what normal operations look like. You’ll also get advanced notice about potential issues as they start to build up and that will give you a better ability to find and fix issues before they become big problems.
Our software, FrameFlow Server Monitor, can do all of the above of course. If you’re new to FrameFlow, take our 30-day trial edition for a spin. If you’ve already purchased a license, book a call with us and one of our reps will help you to implement everything we’ve described here.
Your servers, switches and routers are the core of your operations so of course it’s important to ensure that everything is running smoothly and to get early warning about potential problems. With this set of five best practices you can optimize your server monitoring configuration for easy management, quick adaptation to changing conditions and early detection of potential problems.
Make sure that you have core monitoring configured for all of your systems. For servers you want to be collecting CPU usage, drive space, memory usage and bandwidth. For switches and routers you want to use SNMP monitoring to collect bandwidth levels on each interface. Having core monitoring in place 24×7 will give you clear views of system performance allowing you to detect potential issues.
Each of your systems has its own individual role but groups of systems have shared properties too. Your server monitoring configuration should take advantage of this to the maximum extent because it will save you time when new systems are added, or configuration changes are required. Create authentication profiles and use those for monitoring. Then if credentials are changed, you can update the profile and all of your monitoring will pick up the change automatically.
Dig deeper into profiles by doing the same thing for notifications. You probably have three or more classes of alerts. At the lowest level are informational alerts for items like login events, then there are warnings for issues like low disk space, finally there are critical alerts for when vital systems are down. Define notification profiles for each of these cases and then assign them accordingly. If you ever need to change who is alerted or when they are alerted, you can just modify the profile and all monitoring actions will pick up the change automatically.
Build dashboards that show the status of your criticial IT systems. You can start with the default dashboards that come with our software and customize them by resizing existing elements, adding new ones and designing network diagrams. Then display your dashboards in full screen mode on a large screen in your network operations room so your entire staff has an at-a-glance view of the monitoring status.
Configure one or more reports to build on and deliver them to your inbox on a weekly basis. While you’re busy with important IT issues, it’s easy to neglect your monitoring configuration. Scheduled report delivery is a great way to remind you to take a look at recent results and it helps you to spot trends.
These best practices for server and network monitoring will help you create a solid monitoring configuration that is easy to maintain and provides all of the core data points that you need to stay on top of IT operations. With these basics in place, you can then build out more monitoring to cover specific issues and special cases. You’ll find that your monitoring configuration will evolve over time as new systems, services, applications and hardware come online, but you’ll always be on top of things and able to respond to emerging issues.
If you’re interested in putting these server monitoring best practices in play, you can try our remote server administration tools and monitoring software for 30 days (no costs or obligations).