3 ms·
I've been encountering a similar memory leak on Windows, so I don't know if it's really related to Linux. What I do know is that the way Linux GUI distros deal
by jeroenhd 6y ago
I've been encountering a similar memory leak on Windows, so I don't know if it's really related to Linux.
What I do know is that the way Linux GUI distros deal with low memory situations is absolutely garbage. The new systemd OOM daemon that's shipping with systemd 248 will hopefully improve this situation, but for now I'm left running nohang[1] on my dev machine, where I'm consistently running out of RAM (IntelliJ + tons of Java frameworks + huge React code base + a web browser all take up way too much space!). Enabling zRAM also seems to work, but I haven't measured the effectiveness of that yet. On my home PC with double the ram (32GiB) everything runs smoothly, but the way Linux on the desktop deals with OOM situations is still atrocious. Besides, none of these tools should be consuming such ungodly amounts of RAM anyway.
For server situations, Linux handles OOMs better than Windows, IMO. On desktop, the same strategy just doesn't work and only makes working on it more painful.
[1]: https://github.com/hakavlad/nohang https://github.com/hakavlad/nohang
- bmn__ 6y agoAs of now, I find `earlyoom` is better choice than `nohang` for three reasons: 1. It comes preconfigured: https://build.opensuse.org/package/view_file/openSUSE:Factory/earlyoom/earlyoom.sysconfig https://build.opensuse.org/package/view_file/openSUSE:Factor... 2. Packaging adoption: https://repology.org/project/earlyoom https://repology.org/project/earlyoom https://repology.org/project/nohang https://repology.org/project/nohang 3. Written in C and unburdened with tangential features like desktop notifications, the daemon's process takes very little amount of memory.
- chaimanmeow 6y agoyes, it isn't because of the linix. this is just the mozilla open sores; they will always add a memory vulnerability.