Re: pkg upgrade, vm.pageout_oom_seq: swap off: UFS, ZFS

On Sep 4, 2025, at 00:17, Graham Perrin <grahamperrin@gmail.com> wrote:

> On 04/09/2025 01:53, Mark Millard wrote:
> 
>> On Sep 3, 2025, at 17:14, Graham Perrin <grahamperrin@gmail.com> wrote:
>> 
>>> …
>>> 
>>> Here, with 3 G, fast storage, and swap off: UFS and ZFS seemed much the same.
>>> 
>>> RELEASE, quarterly.
>>> 
>>> …
>> You do not report observations of how large the
>> Swap Used got to be …
> 
> 
> "swap off" in the subject line, and above ;-)

Which is what limits the information about how close
your were to having problems when it did not OOM:
Ifr it was just barely working, little Swap Space
left, that is rather different than if SWAP was
nearly unused, for example. That "swap off" choice
is what I was referencing.

But I did not write enough for my reference to make
any sense. Thus my later note that did try to indicate
why for what would be an example of monitoring "swap
on"  behavior (Swap Used sampling).

> I began focusing on fast storage after realising that with slow storage, there seems to be a massive risk reduction with
> vm.pageout_oom_seq=120

Yep.


===
Mark Millard
marklmi at yahoo.com