100% CPU usage on Linux with v2.1.6 even when syncing is completed

Two Debian boxes auto-upgraded to v2.1.6 today and Syncthing was at 100% CPU usage for several hours. The GUI shows Up to Date for all folders and servers and download and upload rates are at zero. Downgrading to v2.1.5 fixed the issue and they’re both happily idling again.

This didn’t happen on an Ubuntu server or any of the Windows servers in the cluster. They’re running v2.1.6 and the CPU is idling as usual.

I didn’t see anything in the logs when the servers were running at 100% CPU, so I’m not sure what the processes were doing to cause the high usage. Upgrading to v2.1.6 again reproduced the problem on both of them.

A support bundle would help us diagnose the issue if you’re willing to upgrade one of your nodes again.

Web UI → Gear icon top right → Support Bundle

I can see a situation where this might happen when there are folders that haven’t had any activity (changed files) in a long time. (Or if there are database inconsistencies, but I’m not sure what those would be or how they would occur.)

If you feel like it, you could zip up the index-v2 dir and send to me and I could perhaps draw a more confident conclusion, but even without that I have a build that should hopefully resolve the issue; please test and report back :folded_hands:

https://cdn.kastelo.net/tfi/syncthing-2.1.7-dev.2.g831138fb-idxfinalhole-amd64.deb

https://cdn.kastelo.net/tfi/syncthing-2.1.7-dev.2.g831138fb-idxfinalhole-arm64.deb

The v2.1.7 version you provided resolved the issue. CPU usage is back to what it used to be.

Here’s a tarball of index-v2: index-v2.tgz - pCloud

And this is the support bundle from when it had high CPU: support-bundle-JPSS3XO-2026-10-07T034605.zip - pCloud

Nice, thank you; that confirms both the workaround and in fact clarifies the root cause of the issue as well.

I had the same problem with 2.1.6 on Debian. Additional to 100% CPU usage, the process died with an OutOfMemory.

Downgrading to 2.1.5 fixed it as well.