>legacy BIOS
Stop using the L-word, please: BIOS is not "legacy", it's just alternative
way (one of many). All my laptops use BIOS, by the way.
ср, 1 июл. 2020 г. в 18:26, Leigh Scott :
> Has UEFI support improved in the last 2 - 3 years?
> My last experience with it was so poor on a multi-boot
>If both RAM and swap go below 10% free, earlyoom issues SIGTERM to the
process with the largest oom_score. If both RAM and swap go below 5% free,
earlyoom issues SIGKILL
Fedora's earlyoom package is provided with the changed default settings:
```
EARLYOOM_ARGS="-r 0 -m 4 -M 409600 --prefer '^Web
>By killing their software, while it's legitimately using RAM, as expected?
Memory usage causes system hang is not legitimately using RAM. Expected
behavior is killing memory hog.
чт, 9 июл. 2020 г. в 23:53, John M. Harris Jr :
> On Thursday, July 9, 2020 6:24:41 AM MST Sergio Belkin wrote:
> >
>You can limit the process with cgroups
In this case the process will be killed by OOM killer. Note that OOM killer
exists by default. Limiting cgroup means the OOM killer will be invoked in
the cgroup. With earlyoom you database server will get SIGTERM and maybe
will gracefully shutdown.
пт, 10
>75% and 6G
>for the cap might be better.
I agree. I think that 4GB cap is too small and 4 GB may be used quickly.
>I regularly see 3 to 1 and even 4 to 1.
It's true. It's easy to get 4:1 with browser. In fact, the compression
ratio is highly dependent on the workload. I get 1.4:1 with blender,
>oomd (or rather systemd-oomd) and will make heavy use of cgroups to work
well
Other side of this is high CPU usage:
https://github.com/facebookincubator/oomd/issues/79
сб, 11 июл. 2020 г. в 17:34, Benjamin Berg :
> On Sat, 2020-07-11 at 08:43 +0200, Igor Raits wrote:
> > On Fri, 2020-07-10 at 1
> MemAvailable is a bad heuristic, it does *not* represent "free"
> memory.
MemAvailable is a good heuristic, it does represent available memory.
вт, 14 июл. 2020 г. в 00:49, Benjamin Berg :
> On Mon, 2020-07-13 at 08:21 -0700, John M. Harris Jr wrote:
> > On Monday, July 13, 2020 1:58:30 AM MS
> Upon reading the SwapOnZRAM feature proposal, I see it is advocating
> allocating 50% of ram for swap
50% means zram disksize (max size of uncompressed data stored in zram
device) = 50% MemTotal. This data will be compressed, so, its size in RAM
will be smaller than 50% MemTotal (maybe 25% with
> The most
> common value I've found over a long period of time, for swap without
> hibernation is 50% of RAM.
With low RAM (2G) it's easy to use swap on zram with disksize = 150%
MemTotal with opening browsers.
50% maybe OK with MemTotal=8G.
> I'd like to hear from Alexey what he thinks about f
> And this will turn bad quickly, as it will eventually
> include the executables/libraries that must be loaded as they are doing
> work!
> So, we don't want to get the kernel into the situation where it must
> remove executables/libraries from main memory. If that happens, you can
> end up hittin
>Killing users' programs needlessly is not welcome
Setting limits for cgroup (MemoryMax, MemorySwapMax) leads to "Killing
users' programs needlessly": system-wide available memory may be not
exhausted, but OOM killer will be invoked in this cgroup.
>The goal is to ensure the kernel can keep doing
> Userspace isn't dead
If something looks like dead it is dead. A userspace that does not respond
to user actions is a dead userspace.
вс, 19 июл. 2020 г. в 17:54, John M. Harris Jr :
> On Sunday, July 19, 2020 1:47:23 AM MST Alexey A. wrote:
> > >Killing users' p
> may be different for GNOME with JIT'ed code
> potentially getting swapped out
How to recognize JIT'ed code? Do you mean that we shouldn't lock it?
пн, 14 сент. 2020 г. в 20:17, Benjamin Berg :
> On Sun, 2020-09-13 at 16:47 +, Alexey Avramov wrote:
> > > So with
> > > applications started,
> memory is executable code, but it behaves like
> other process memory such as the heap or the stack
Seems like it is not a problem: with GNOME effect is the same as with other
environments.
пн, 14 сент. 2020 г. в 21:14, Benjamin Berg :
> On Mon, 2020-09-14 at 20:35 +0900, Alexey
> system hangs with earlyoom and swap on zram, disksize=150% MemTotal,
compression ratio at the end is 1.4, SwapFree=25%, the zram device occupies
80% of the memory with running Blender https://imgur.com/a/wNvBVIn.
> With disksize=100% MemTotal this should be very rare case.
-- https://pagure.io/
> Fedora is. Linux isn't. Linux is just an OS kernel
Linux (/ˈlinʊks/ (listen) LEEN-uuks or /ˈlɪnʊks/ LIN-uuks[9]) is a
family of open-source Unix-like operating systems based on the Linux
kernel.
ср, 10 мар. 2021 г. в 01:35, Vitaly Zaitsev via devel
:
>
> On 09.03.2021 16:43, Matthew Miller wrot
What is in the journal?
пн, 22 мар. 2021 г. в 03:12, Vascom :
>
> Hi all.
>
> I want to draw your attention to the bug in systemd-oomd in F34.
> This oomd kills my work application during the second run of the stress test.
> And I create this ticket:
> https://bugzilla.redhat.com/show_bug.cgi?id=1
The bug https://bugzilla.redhat.com/show_bug.cgi?id=1941340
пн, 22 мар. 2021 г. в 03:12, Vascom :
>
> Hi all.
>
> I want to draw your attention to the bug in systemd-oomd in F34.
> This oomd kills my work application during the second run of the stress test.
> And I create this ticket:
> https://b
Hi,
Userspace OOM killer nohang [1] is orphaned [2] now.
This oom killer supports PSI and GUI notifications.
The package may be of interest to users who are not suitable for
systemd-oomd for some reason. For example, systemd-oomd may kill the whole
session (this applies to XFCE, Mate, LXDE etc.
related: systemd-oomd kills the whole session if it's started from
console https://bugzilla.redhat.com/show_bug.cgi?id=1933494
ср, 28 июл. 2021 г. в 10:28, David Airlie :
>
> Hi,
>
> I've been using f34 with gnome/wayland session on a 16MB machine which
> has 8GB zram swap configured.
>
> If I bui
20 matches
Mail list logo
