# Which log level to use to debug poor Main UI performance (e.g. 30 sec to load Semantic Model)

**URL:** <https://community.openhab.org/t/which-log-level-to-use-to-debug-poor-main-ui-performance-e-g-30-sec-to-load-semantic-model/145100>\
**Category:** Runtime\
**Tags:** delay, performance, loading, semantic-model, responsiveness\
**Created:** [March 9, 2023, 9:54pm UTC](https://community.openhab.org/t/which-log-level-to-use-to-debug-poor-main-ui-performance-e-g-30-sec-to-load-semantic-model/145100 "2023-03-09T21:54:14Z")\
**Posts on this page:** 1\
**Showing post:** 10

<div class="post-metadata">

**Author:** ![Cplant](https://community.openhab.org/user_avatar/community.openhab.org/cplant/32/90960_2.png) [@Cplant](https://community.openhab.org/u/Cplant)\
**Post date:** [March 20, 2023, 7:44pm UTC](https://community.openhab.org/t/which-log-level-to-use-to-debug-poor-main-ui-performance-e-g-30-sec-to-load-semantic-model/145100/10 "2023-03-20T19:44:44Z")

</div>

> [@mstormi](#):
>
> You should be moving to openHABian, it does swapping (well, _paging_ it actually is) in ZRAM, i.e. keeps the paged-out pages in RAM but compresses them.

> [@rlkoshak](#):
>
> Don’t use Docker which consumes more RAM. […] Of course, the easiest would be to use openHABian which does a whole lot to optimize the RAM usage as well as preserve the SD card by limiting writes.

Yes, I know. However, there are two reasons for being on Docker (and I also remember your post [here](https://community.openhab.org/t/best-way-to-build-a-new-openhab-install-today/144308/4) @rlkoshak, which was spot on). First, I did not get [wmbusmeters](https://github.com/wmbusmeters/wmbusmeters) running in the first place on openHABian when I started with openHAB two years back (but it could’ve also been my mistake), and got it running immediately with the wmbusmeters docker image, and second, well, what shall I say, this entire “Docker/Portainer thing” is just my guilty pleasure as I wanted to learn some new piece of tech, …, and now I’m paying the price with a high RAM load and the need to buy new Hardware because of that. 😉

> [@wborn](#):
>
> Why do you have 7 `influxd` processes each consuming memory? I only have one. 🙂

Good question. II assume I did a [pretty standard docker install](https://community.openhab.org/t/install-maintain-openhab-on-raspi-in-docker-together-with-mosquitto-duplicati-influxdb-samba-etc-step-by-step-instruction/141689). Any idea how I can check why this is the case? Besides the strange entries in `htop`, I do not experience any other problem.

---

_[View the full topic](https://community.openhab.org/t/which-log-level-to-use-to-debug-poor-main-ui-performance-e-g-30-sec-to-load-semantic-model/145100)._
