RAM Size Guidance for Dragonfish

[winnielinnie]

@Captain_Morgan

We expect to be able to recommend best mitigation and plans for a fix, on Friday.

If you come back with “We’ve decided it’s best to recommend new installations to have a minimum of 512 GiB of RAM”, then I’m deleting my account.

[essinghigh]

Sounds like a perfect solution, along with limiting ARC to 50% so everyone has plenty of memory free!

[somethingweird]

@winnielinnie

“We’ve decided it’s best to recommend new installations to have a minimum of 512 GiB of RAM”

Give me a good excuse to tell Mrs. Potatohead - I need more ram to run the next version of truenas. Yes - I need to upgrade to the next version for security updates!

[winnielinnie]

I’m sorry, but I just have to drop this in here.

I didn’t say it! The A.I. chatbot said it!

[victor]

I had to chuckle…

Seriously, I’m assuming that for Linux it involves far more than simply setting “total RAM minus 1GB” to get it to work?

[Captain_Morgan]

@winnielinnie

If you come back with “We’ve decided it’s best to recommend new installations to have a minimum of 512 GiB of RAM”, then I’m deleting my account.

Ha! No expectations that any hardware changes are required…

[somethingweird]

Are you trying to get ix to drop SCALE - and keep going with CORE?!

[winnielinnie]

No.

Yes.

[kris]

Cross-posting for reference:

Investigation of slow UI / RAM / SWAP issues

If you’ve been on the forums this week you may have noticed a few threads about various issues with UI responsiveness and swap/memory issues on 24.04.0. (Dragonfish). We’re still in the background working on a proper fix for 24.04.1 (due in a few weeks). The underlying issue is rather tricky with a lot of nuance. We’ll share more details on the technical side soon, don’t want to steer folks too wrong until we are 100% sure what the final patch will look like. In the meantime, if you find your s…

[patrickkeane]

Cross-posting for visibility.

We can confirm that the significant contributor to the issue that has been reported in this thread and others is related to a Linux kernel change in 6.6. While swap adjustments proposed can mitigate the issue to some degree, the multi-gen LRU code in question will be disabled by default in a subsequent Dragonfish patch release, and should resolve the problem. This can be done in advance following the directions below from @mav. Thanks to all of the community members who helped by reporting the issue which aided in reaching the root cause. Additional feedback is welcome from those who apply the change below.

@mav

After deeper digging into the problem, it looks to me caused not only by increased ARC size, reduction which according to some people may not fix the problem, but also by Linux kernel update to 6.6, which enabled previously disabled Multi-Gen LRU code, written in a way that assumes that memory can not be significantly used by anything other than page cache (which does not include ZFS). My early tests show that disabling it as it was in Cobia with echo n >/sys/kernel/mm/lru_gen/enabled may fix the problem without disabling swap.