How to Identify and Resolve Server Bottlenecks in Your FileMaker Application to Encourage Scaling

In this fifth post in our Scaling FileMaker blog series, we’re sharing how to identify the source of the slowdowns in your application and make big improvements to your system to promote a better user experience and encourage your business to scale.

There are a variety of places to look when facing a server bottleneck and performance issues in your FileMaker application. After managing this process dozens of times for clients of all sizes and types, we’ve put together a step-by-step process aligned with our best practices:

1. Review Your Data

Performance Log in FileMaker Server

The real first step here is to make sure you’re logging call stats in FileMaker Server. It’s not a default setting, so make sure it’s turned on and set up properly.

Once you’re collecting metrics from FileMaker Server, it’s time to start digging into your client stats and call stats. What sticks out to you as unusual?

Operating System and Server Data

Then, review metrics from your operating system and the server itself. (Soliant.cloud clients, these would be your Windows Server metrics.)

We recommend you look at the following data:

• CPU utilization
• Disc utilization
• Network throughput

Plus, any other advanced data you have access to.

Anecdotal Information

Don’t forget to talk to your users about what they’re experiencing and when it typically happens. Ask a lot of questions and take good notes. You can then compare these situations to your application scripts and processes.

2. Replicate the Issue

Once you’ve gotten some direction from users on when and where the server bottlenecks happen, it’s time to experience the performance issue yourself. If you can replicate it yourself, you not only understand the problem better but can also collect remote call logs. These are the calls made between FileMaker Pro and Server.

The remote call log file indicates the type of call, how many calls are made, and how long it takes for a call to happen. For example, if you click a button and the expected action takes a long time to finish, you should look at remote call logs to see why. In one client’s situation, we realized one button made five thousand calls with every click. Ideally, we aim for fewer than five calls total. We resolved it, and performance was significantly improved immediately.

I had two sessions on this topic at AutoEnter Live in July 2022. If you want to take a deep dive into this topic, you can do so via the recordings on YouTube:

1. Part 1: Visualizing Network Communication Between FMP/FMGo and FMS
2. Part 2: Visualizing Network Communication Between FMP/FMGo and FMS

These sessions clarify why remote call logs are so crucial. You can make many significant changes to your performance by the end of this step. Our team unclogs most bottlenecks at this point.

3. Advanced Investigation

If you can’t replicate the issue or need to look deeper within your data to untangle a more complex server bottleneck, it’s time to move into more advanced tactics.

Overlaying Data

Above, we shared the importance of looking at both your FileMaker Server and operating system data. However, if you need to take a more comprehensive look, we recommend you overlay the data to see what’s happening all at once, at any given moment in time.

Here’s where you can look for metrics that stand out. For example, one of our clients is struggling with a slow system every morning at 9 am. We used a data overlay to look at a snapshot of all data at that time and found the CPU to be very high, and the system problem was further exacerbated by another background process taking place.

This Root Cause Analysis (RCA) allows us to take a deep look at everything happening on the server and find out what’s causing the performance issue.

Server Upgrading & Overprovisioning

You can also review potential over-provisioning of the server. Many FileMaker administrators who experience slowdowns jump straight to upgrading their servers. However, this is expensive and often doesn’t make a difference. And then they forget to roll back the upgrade. It’s a never-ending cycle we see often.

Our team looks for over-provisioning and rolls back upgrades our clients don’t need to reduce costs during our investigation of server bottlenecks. This often creates a nice offset for the investment a client makes with our team and our services.

4. Testing Fixes & Implementing Updates

At this point, you’ve likely identified multiple potential sources of the slowdown. It’s time to start acting on them.

For example, if you find queries on a specific feed are very slow, it’s time to test several fixes and determine which works best. Or, if your Server schedule is slow, you can isolate metrics related to that script and find the cause of the slowdown. If you’re working with a server hardware bottleneck, this analysis informs you on how much you need to upgrade and the most efficient way to get there.

If you’ve uncovered too many potential issues, we recommend moving forward with the 80/20 rule (also known more formally as the Pareto Principle). Target improving the top 20% of your issues. In most cases, this is enough to make an impact. You don’t need to fix everything, just enough to make a big impact.

Having a diversely talented and skilled team can make a difference, too. At Soliant, we have a full bench of people who can tackle these issues. We divide and conquer to get answers to clients quickly. If you don’t have the extra resources to complete these investigated deep dives, we can help.

5. Documentation

Of course, anytime you uncover a significant finding and make a change (even if it doesn’t work), you need to document it. These notes will serve you well in the future when you need to reference what you’ve already tried if a new problem (or the same one!) pops up again.

Working with a Partner

Our team built our reputation as a top Claris Partner with the innovative new applications we’ve developed for our clients, but over the years, we’ve sustained this reputation by helping organizations take FileMaker further, empowering them to achieve more on the platform.

When our clients face performance issues and server blocks, we aim to get them an answer and a solution as soon as possible. We can often resolve issues without touching code in a short amount of time in a way that makes a big impact for our clients. If you’re facing challenges with your FileMaker application, we’re happy to work with you to investigate why. Contact us to talk with a consultant today.