📦 A new release has been made for this pull request.

To play around with this PR, pull codetheweb/muse:pr-967 or codetheweb/muse:84ad6112dd3b09bb5ff050403327568ae8afc6c4.

Images are available for x86_64 and ARM64.

Latest commit: 84ad611

is this in the masterbuild? I notice in unraid my ram usage just slowly keeps creeping up the longer I listen to the bot.

This was superseded by #976. It looks like the issue should be fixed when running in Docker, but may not yet be fixed if you're running it manually from the repo. How do you run Muse?

Its from the unraid docker. I will give it a few more tests when I have some time.

Unraid CA for muse is using the official codetheweb/muse docker image, just FYI.

As an aside, @GitGurky, this could be due to node's tendency to hoard ram even if its not used. If your unraid server has a ton of free ram and you have not limited it using Extra Parameters in the CA template for muse there is a possibility node is just hanging on to ram because there is more to spare. I don't have an article handy to back this up but I have encountered similar behavior in other apps I develop.

You can force node to try to stay within a specific total ram allotment by passing the node arg

node --max_old_space_size=NUMBER ...

where NUMBER is the megabytes of ram it should try to stay under. This should be compatible with passing to tsx as well.