I never could get a clear understanding of these. I just throw a partition on the nvme for swap as a failover. If it starts to fill, treat it like the countdown timer in uplink.
I got corrected some time ago about this, it seems it’s possible, but not advisable. Like, if zram started, zswap will disable itself.
If you want to test zswap, you could create a swap file, disable systemds zram service and put the zswap variables into the kernel variables, also checking if swappiness is at the normal values, instead of the tuned ones for zram.
In my testing, zram has much, much better compression than zswap.
The points about LRU inversion, cgroups, and so on are valid, but at the end of the day, I don’t really care. I was able to open as many firefox tabs as I wanted with zram, but I could not do so with zswap, and that’s what matters to me.
The author of a blogpost is a facebook engineer. Millions of ultra high performance Linux servers are a very different usecase than a single desktop. It’s perfectly reasonable for a solution for one to not be appropriate for the other.
What a load of bull. Zram is so much more useful. That guy does not even know you can set up a fallback device for zram.
Don’t trust any fool with a blog.
Probably written and “researched” with ChatGPT.
Just switched to zswap on my cachyOS install and I can feel the difference. I had freezes where the system chugs when I have several processes running. Now, I don’t get them anymore. Not gonna pretend that I understand everything about them fully, but I searched online and some folks recommended zswap and I went with it.
bookwyr.me
Share on Mastodon
Share on Reddit
Share on WhatsApp
/swapfile
sudo swapon /deeznutsis my favourite as it’s a valid commandYou have to fallocate /deeznuts first.
or dd(eeznuts) it!
I never could get a clear understanding of these. I just throw a partition on the nvme for swap as a failover. If it starts to fill, treat it like the countdown timer in uplink.
upvote for uplink reference.
Is it possible to have both? I’m finding zram works well until it doesn’t.
I got corrected some time ago about this, it seems it’s possible, but not advisable. Like, if zram started, zswap will disable itself.
If you want to test zswap, you could create a swap file, disable systemds zram service and put the zswap variables into the kernel variables, also checking if swappiness is at the normal values, instead of the tuned ones for zram.
Just to be super nerdy, here’s an amazingly well-written article on the subject from someone with a lot of credibility: https://chrisdown.name/2026/03/24/zswap-vs-zram-when-to-use-what.html
tldr; prefer zswap
In my testing, zram has much, much better compression than zswap.
The points about LRU inversion, cgroups, and so on are valid, but at the end of the day, I don’t really care. I was able to open as many firefox tabs as I wanted with zram, but I could not do so with zswap, and that’s what matters to me.
The author of a blogpost is a facebook engineer. Millions of ultra high performance Linux servers are a very different usecase than a single desktop. It’s perfectly reasonable for a solution for one to not be appropriate for the other.
i prefer ram
What a load of bull. Zram is so much more useful. That guy does not even know you can set up a fallback device for zram.
Don’t trust any fool with a blog.
Probably written and “researched” with ChatGPT.
“That guy” works full time as a kernel developer on the Linux memory management.
You should read the article. It has a section specifically on using a zram fallback device and how its performance compares to a similar zswap setup.
Just switched to zswap on my cachyOS install and I can feel the difference. I had freezes where the system chugs when I have several processes running. Now, I don’t get them anymore. Not gonna pretend that I understand everything about them fully, but I searched online and some folks recommended zswap and I went with it.
Can I make it z-any more obvious?
Computers
magnets
theres zswap?