# Profiling personal minimal instance (unexpected high cum. CPU load) **URL:** \ **Category:** General\ **Created:** [April 27, 2026, 10:25am UTC](https://forum.gitea.com/t/profiling-personal-minimal-instance-unexpected-high-cum-cpu-load/12074 "2026-04-27T10:25:41Z")\ **Posts on this page:** 1\ **Page:** 1
**Author:** ![xor-gate](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.gitea.com/xor-gate/32/6155_2.png) [@xor-gate](https://forum.gitea.com/u/xor-gate)\ **Post date:** [April 27, 2026, 10:25am UTC](https://forum.gitea.com/t/profiling-personal-minimal-instance-unexpected-high-cum-cpu-load/12074/1 "2026-04-27T10:25:41Z")
Hi all, I run a Gitea instance at work and at home. But because my personal NAS/server is a low-end server I like to see where all CPU time is going to. RAM is not a big problem. Here are my instance details: - FreeBSD 15 host OS (16GB DDR3, 4 core intel atom Intel® Atom® Processor C2550) - Gitea v1.25.5 AMD64 in a Jail (container) - sqlite3 DB config (for simplicity) - on ZFS filesystem (single disk for now…) - Git version 2.52.0 After some maintaince I saw it polls every 8 hours for few (non-existing) Github mirrors. I also know Gitea does some scheduled/background tasks for maintainance. I have found under admin panel → Monitoring → Trace to create some trace files. Not sure if they are pprof to create burndown chart diagram (dat files). Here are some stats of FreeBSD `top` ![Screenshot 2026-04-27 at 12.22.14](https://us1.discourse-cdn.com/flex020/uploads/gitea/original/2X/6/6bde94fda5c7f999db6ce653157e6673935fe99b.png) The CPU time of 717:38 is uptime “up 1 day, 2:51”. It feels a little excessive for the small instance with only two users and 14 git repositories. If it is needed I can upload a gitea-diagnosis zipfile of 60 seconds monitoring trace. I did not looked deep into the trace files, but I know sqlite can be a bit more intensive then a “real” database like Postgres. But it would probably be more hungry (for RAM caches).