Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 21 Jan 2019 18:31:53 +0100
From:      Andrea Venturoli <ml@netfence.it>
To:        freebsd-questions@freebsd.org
Subject:   Re: The mysterious kern.maxswzone
Message-ID:  <06a5651b-fdd8-1dcf-d0b1-1e3d4dafe11d@netfence.it>
In-Reply-To: <20190121160302.0a5e2805@gumby.homeunix.com>
References:  <20190116022046.GA45024@admin.sibptus.ru> <50f1f8d5-7db6-4abb-6beb-c82f17396304@netfence.it> <20190121160302.0a5e2805@gumby.homeunix.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On 1/21/19 5:03 PM, RW via freebsd-questions wrote:

> Do you know, for sure, that there isn't a risk of a panic if you run out
> of zone space before you run out of swap.

Short answer: NO!
I don't know for sure, but I think this isn't the case.
 From what I read around, IIUC, zone space is used to manage swap space, 
so the only effect is that some swap space won't be used.
I hope someone with more solid knowledge steps up...



> In amd64 you can't
> increase the zone size because that computed default is also the limit.

Then, why do I get the same warning ("warning: increase kern.maxswzone 
or reduce amount of swap") if the suggested operation is not possible?
I guess the message should be patched.



> The value is in bytes, so it isn't huge. Try setting it to a few percent
> above
> 
>    36175872 * 131072 / 113792
> 
> something like 42500000

"Huge" compared to what? :)
The OP has 512MiB of RAM: 42500000 byte is not a negligible amount in 
that case, but he is the only one who knows if it's worth, based on his 
load.

  bye
	av.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?06a5651b-fdd8-1dcf-d0b1-1e3d4dafe11d>