Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 15 Nov 2019 15:24:00 -0800
From:      Xin LI <delphij@gmail.com>
To:        Eugene Grosbein <eugen@grosbein.net>
Cc:        FreeBSD stable <freebsd-stable@freebsd.org>, Alexander Motin <mav@freebsd.org>
Subject:   Re: bhyve memory leak in stable/11
Message-ID:  <CAGMYy3txqYg34UBQeLToSN-Thsfp0ZuBOuWTaPHS8VMrhe-Szg@mail.gmail.com>
In-Reply-To: <7fddcea5-2188-afe1-3ea9-a53dffdbec32@grosbein.net>
References:  <7fddcea5-2188-afe1-3ea9-a53dffdbec32@grosbein.net>

next in thread | previous in thread | raw e-mail | index | archive | help
There are several leaks fixed recently, but not all of them were merged
back to stable/11 (e.g. r341705 which may or may not be related to your
issue).

Personally, I'd recommend updating to 12.x for new features as they are
being actively developed and would be supported for about 3 more years
compared to 11.x series.

On Thu, Nov 14, 2019 at 7:25 AM Eugene Grosbein <eugen@grosbein.net> wrote:

> Hi!
>
> After 1 day and 11 hours of host's uptime I observe significant leak of
> bhyve instances RSS:
>
>   PID USERNAME    THR PRI NICE   SIZE    RES STATE   C   TIME    WCPU
> COMMAND
> 66668 root         34  20    0 60320M 58866M kqread 21  62.0H 692.35%
> bhyve: sappdev (bhyve)
>  5546 root         26  20    0 22573M 22088M kqread 12 310:39   6.41%
> bhyve: srdapp01 (bhyve)
>  3842 root         20  20    0 15946M 15862M kqread 13 340:36  14.03%
> bhyve: sdc01 (bhyve)
>  4729 root         20  20    0 13792M 13455M kqread  8  79:40   1.02%
> bhyve: skms01 (bhyve)
>  4283 root         28  20    0 12721M 12682M kqread 29  92:23   0.80%
> bhyve: sfile01 (bhyve)
>  8057 root         34  20    0  8512M  8183M kqread 25 186:52  17.02%
> bhyve: ssql01 (bhyve)
> 23336 root         34  20    0 16418M  7716M kqread 14 114:23  17.62%
> bhyve: solap (bhyve)
>
> The package vm-bhyve-1.2.3 is used here to manage bhyve instances
> and sappdev.conf has memory=40G limit. srdapp01 is limited to 12G,
> sdc01 to 4G, skms01 to 2G, sfile01 to 2G, ssql01 to 8G and solap to 16G.
>
> Most of are using much more memory than they should. All of them run
> different versions of Windows servers.
>
> The host is 11.3-STABLE/amd64 r354667 with 360G RAM and guest systems use
> ZFS ZVOLs and vfs.zfs.arc_max=160g.
>
> How do I debug this?
> _______________________________________________
> freebsd-stable@freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/freebsd-stable
> To unsubscribe, send any mail to "freebsd-stable-unsubscribe@freebsd.org"
>



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAGMYy3txqYg34UBQeLToSN-Thsfp0ZuBOuWTaPHS8VMrhe-Szg>