If the spark profiler is not updating, the problem usually comes from one of a few things. The live viewer may still be within its normal refresh window, the profile may not have collected matching data, or your spark setup may be using an older version or a command with filters that are too strict.
This guide explains why spark profiler stops refreshing, why the viewer may show no data, and what to do step by step to get it working again. It starts with the easiest checks first and then moves to version, browser, and environment fixes.
Why Is Spark Profiler Not Updating?
The spark live viewer is not meant to refresh every second. The maintainer says top statistics refresh every 10 seconds, while full profiling data refreshes every minute. That means the page can look stuck when it is still working normally. There are also cases where the profile does not collect useful samples, especially if the command uses strict filters or the server is too idle.
Main causes include:
- You Did Not Wait for the Normal Refresh Window
- The Profiler Collected No Matching Data
- Your Command Used Filters That Were Too Strict
- The Viewer Opened an Old or Stale Report
- Your spark Version Is Outdated
- Java or Platform Compatibility Is Causing Problems
- The Viewer Upload or Open Step Failed
How to Fix Spark Profiler Not Updating
Follow the fixes below one by one until the problem is solved. Start with the simplest checks first, then move to filters, version updates, and environment issues if the viewer still is not refreshing.
1. Wait for the Normal Live Refresh Cycle
This is the easiest check and the one most people skip. spark’s live viewer does not show every sample instantly. The maintainer says statistics refresh every 10 seconds and full profiling data refreshes every minute.
So if you just started profiling, give it a little time before assuming the viewer is broken.
A simple rule helps:
- Wait at least 10 seconds for summary changes
- Wait up to 1 minute for deeper profiling data
- Keep the profiler running during real activity or lag
2. Rerun the Profiler Without Strict Filters
Filters can make a profile look empty even when spark itself is working. A current GitHub issue shows that a command using –only-ticks-over 100 –timeout 60 produced a profile with no data in at least some setups.
If your current command is strict, strip it back and test again.
Try a simpler run first:
- Start with /spark profiler –timeout 60
- Remove –only-ticks-over
- Remove extra filters for the first test
- Profile while the server is under real load
Once the viewer starts updating normally, add your filters back one at a time.
3. Use /spark profiler open Again
The live viewer depends on opening the running profile correctly. spark introduced the live viewer flow around /spark profiler open, so if the viewer page is stale, reopening it can help.
There is also a reported issue where /spark profiler open produced an error in some environments instead of opening properly.
If the page looks frozen:
- Run the profiler again
- Use /spark profiler open
- Make sure you open the fresh link, not an older saved one
- Check that you are viewing the active profile session
4. Refresh the Browser Tab
Sometimes the problem is not spark itself. The viewer tab may simply be stale.
Try these steps:
- Refresh the Tab
- Reopen the Viewer Link
- Open the Link in an Incognito Window
- Test in Another Browser
- Close Older spark Viewer Tabs
This helps rule out cached or stale pages before you change anything on the server.
5. Update spark to the Latest Version
Old spark builds can cause viewer problems, upload issues, or compatibility trouble. spark 1.10.119 includes a fix for running on Java 22+ JVMs, so updating matters even more if your server uses a newer Java version.
If you have not updated spark in a while, do that before trying more advanced fixes.
This is especially important if:
- You recently updated Java
- You moved to a newer server build
- The profiler used to work and now does not
- The viewer is opening but not showing useful results
6. Check Java and Server Platform Compatibility
spark runs across different environments such as Paper, Spigot, Fabric, and Forge. A working setup on one platform or Java version may behave differently after a major upgrade.
Check these basics:
- Your Current Java Version
- Your Server Platform
- Your spark Build Version
- Recent Updates to the Server or JVM
If you upgraded Java recently, update spark too. That is one of the easiest ways to avoid hidden compatibility issues.
7. Check for Upload or Network Problems
If spark cannot open or upload the live report properly, the viewer may look broken even when the profiler is running. One reported issue around /spark profiler open included a DNS lookup problem, which shows that connectivity can be part of the failure.
A few quick checks can help:
- Confirm the Server Has Normal Network Access
- Check DNS Resolution
- Test Whether spark Viewer Links Open Properly
- Try Again After Restarting the Server Session
If the open step fails, the update problem may really be an access or upload problem.
8. Try a Clean Basic Profiling Command
A clean test gives you the fastest answer. If a basic profiler run updates normally, the issue is likely in your filters or workflow, not the core tool.
Start with this simple approach:
- Run a plain profile
- Add only a timeout
- Open the live viewer
- Watch for the first refresh cycle
- Add filters later only if needed
This method removes guesswork and makes troubleshooting much easier.
9. Check Whether the Server Was Too Idle
spark needs useful samples to show useful results. If the server or client is mostly idle, the viewer may appear empty or barely updated.
This happens more often than people think. A profiler run during a quiet period can finish with little or no useful data.
Try profiling during:
- A Lag Spike
- Chunk Loading
- Heavy Plugin Activity
- Combat or Mob Activity
- Large Player Movement
- World Generation
That gives the profiler something real to capture.
10. Look for the “No Data” Case Separately
Sometimes the main problem is not that the viewer failed to refresh. The real issue is that the profile contains no matching data.
That can happen because:
- The command filtered out all relevant ticks
- The timeout ended too soon
- The server was too quiet
- A current spark issue affected the profile output
If the viewer says there is no data, simplify the command and run the profiler again during actual load. A recent GitHub issue confirms that this can happen even when the command looks valid.
How to Prevent This Problem from Happening Again
These small steps can reduce the chance of the same issue coming back.
- Keep spark Updated
- Use a Simple Command First
- Avoid Strict Filters at the Start
- Wait for the Normal Refresh Window
- Test the Viewer in a Fresh Browser Tab
- Update spark After Major Java Changes
- Profile During Real Server Activity
- Check Release Notes After Major spark Updates
Final Thoughts
When spark profiler is not updating, the cause is usually simpler than it looks. In most cases, the viewer is still within its normal refresh cycle, the command is too restrictive, or the setup needs a spark update.
Go through the fixes in order and keep the first test simple. Once the live viewer updates again, you can add filters and longer profiling runs with much more confidence.