Date: Wed, 28 Jun 2006 08:40:56 +0800 From: "Ronnel P. Maglasang" <rmaglasang@infoweapons.com> To: Artis Caune <Artis.Caune@latnet.lv> Cc: freebsd-pf@freebsd.org Subject: Re: maximum number of queues Message-ID: <44A1D018.7010003@infoweapons.com> In-Reply-To: <EEB6FF48-0AFE-4F6F-ABAF-7D22C71C476C@latnet.lv> References: <449FC649.4030205@infoweapons.com> <449FE1FE.5060308@bitparts.org> <EEB6FF48-0AFE-4F6F-ABAF-7D22C71C476C@latnet.lv>
next in thread | previous in thread | raw e-mail | index | archive | help
Artis Caune wrote:
>> Ronnel P. Maglasang wrote:
>>> hi, is there a maximum system limit for the number of queues for
>>> altq? is this settable/tunable? whats the default limit?
>>>
>
>
> # grep _MAX_ sys/contrib/altq/altq/*.h
>
> edit, tune && recompile kernel
>
i came across an old thread that answered the same problem, it said
modify the following typedef:
#define HFSC_MAX_CLASSES 64
is this enough? whats the logic behind the changes? is there a formula to
this?
looking at the code, pf seemed to run out of memory when it allocates
an "altq" data from the altq pool.
sys/contrib/pf/net/pf_ioctl.c
--
--
case DIOCADDALTQ: {
struct pfioc_altq *pa = (struct pfioc_altq *)addr;
struct pf_altq *altq, *a;
if (pa->ticket != ticket_altqs_inactive) {
error = EBUSY;
break;
}
altq = pool_get(&pf_altq_pl, PR_NOWAIT);
if (altq == NULL) {
error = ENOMEM;
break;
}
--
--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?44A1D018.7010003>