> On Nov 27, 2017, at 04:33, George Amanakis wrote:
>
>
>> Whatever your primary use case is? My biggest concern is that it
>> simply not crash - 300 second long tests, 1200 seconds, all night
>> long over and over, again, pounding it flat.
>
> My home router runs x86_64 Archlinux on net-next
George Amanakis writes:
>> Whatever your primary use case is? My biggest concern is that it
>> simply not crash - 300 second long tests, 1200 seconds, all night
>> long over and over, again, pounding it flat.
>
> My home router runs x86_64 Archlinux on net-next with cake and nf_conntrack
> compi
On Sun, 26 Nov 2017, Kevin Darbyshire-Bryant wrote:
I don't really know. We want cake to compile without conntrack
(it's expensive).
I’m slightly confused by that statement - The conntrack lookup is controlled
per instance and is one conditional in the code path:
It's not the cake lookup th
Whatever your primary use case is? My biggest concern is that it
simply not crash - 300 second long tests, 1200 seconds, all night
long over and over, again, pounding it flat.
My home router runs x86_64 Archlinux on net-next with cake and
nf_conntrack compiled as integrals. TSO, GSO and GRO
...
} else if (!sch->q.qlen) {
int i;
for (i = 0; i < q->tin_cnt; i++) {
if (q->tins[i].decaying_flow_count) {
qdisc_watchdog_schedule_ns(&q->watchdog, now +
q->tins[i].cparams.target);
I have sat here for a half hour now trying to figure out if the
watchdog can be called when in unlimited mode (e.g., not shaped), and
got nowhere.
--
Dave Täht
CEO, TekLibre, LLC
http://www.teklibre.com
Tel: 1-669-226-2619
___
Cake mailing list
Cake@
It won't link unless conntrack is in the kernel, and *that* is costly for
some.
What we could do is make NAT support optional in Kconfig and have that
option depend on conntrack. Would need a little fettling of cake itself to
make the NAT support properly optional at compile time.
- Jonathan Mor
I just committed this
diff --git a/sch_cake.c b/sch_cake.c
index 14b32e0..606ec29 100644
--- a/sch_cake.c
+++ b/sch_cake.c
@@ -654,6 +654,20 @@ static inline void cake_update_flowkeys(struct
flow_keys *keys,
}
#endif
+/* Cake has several subtle multiple bit settings. In these cases you
+ * wou
> On 26 Nov 2017, at 18:22, Dave Taht wrote:
>
> Toke Høiland-Jørgensen writes:
>
>> gamana...@gmail.com writes:
>>
>>> Just finished building, setting both sch_cake and nf_conntrack as
>>> integral succeeds. Setting nf_conntrack as module fails with sch_cake
>>> as integral (makes sense).
>
On 26 November 2017 19:22:35 CET, Dave Taht wrote:
>Toke Høiland-Jørgensen writes:
>
>> gamana...@gmail.com writes:
>>
>>> Just finished building, setting both sch_cake and nf_conntrack as
>>> integral succeeds. Setting nf_conntrack as module fails with
>sch_cake
>>> as integral (makes sense).
> On Nov 26, 2017, at 7:19 PM, Dave Taht wrote:
>
> Pete Heist mailto:petehe...@gmail.com>> writes:
>
> I think you can safely drop pfifo from future tests.
Yeah, done, chuckled at that actually. I was using pfifo as the “cold pool”.
> I'd rather like combined charts so it is possible to eyeb
> On Nov 26, 2017, at 6:45 PM, Dave Taht wrote:
>
> the real test of a aqm is how it handles a range of rtts in real traffic.
>
> While a simple, useful test is merely to insert delays of say, 20, 40, 80,
> 160ms
> inline with
>
> a real test would have multiple tcp targets of those delays.
>
Toke Høiland-Jørgensen writes:
> gamana...@gmail.com writes:
>
>> Just finished building, setting both sch_cake and nf_conntrack as
>> integral succeeds. Setting nf_conntrack as module fails with sch_cake
>> as integral (makes sense).
>
> So this probably needs either a Kconfig dependency or an i
Kevin Darbyshire-Bryant writes:
>> On 26 Nov 2017, at 10:00, Jonathan Morton wrote:
>>
>> On a purely theoretical basis, it probably isn't as fast, because now the
>> 'found' condition has to be tested multiple times on the fast and slow paths
>> alike. Only if the compiler is smart enough to
Pete Heist writes:
> I have a script (called flenter) which can run flent with different parameter
> variations and produce an html report. I’m very sorry again that this doesn’t
> yet use flent’s batch feature- it was started before I knew about that.
>
> I’ll use it to produce some “rounds” of
On Sun, Nov 26, 2017 at 1:47 AM, Pete Heist wrote:
>
> On Nov 26, 2017, at 9:42 AM, Dave Taht wrote:
>
> That is a really, really long, and extremely pleasant, way of saying:
>
> "OK, it doesn't crash".
>
> :)
>
> can flenter work with the veth stuff and namespaces?
>
>
> Maybe a little more than
gamana...@gmail.com writes:
> Just finished building, setting both sch_cake and nf_conntrack as
> integral succeeds. Setting nf_conntrack as module fails with sch_cake
> as integral (makes sense).
So this probably needs either a Kconfig dependency or an ifdef?
-Toke
_
Have to say, my perception is that I’ve a small cpu usage improvement for my
use case with these changes. Then I would expect that ‘cos I use
‘dual-srchost’ on egress and ‘dual-dsthost’ on ingress, so I’m saving running
the dst/src host allocation code on egress/ingress respectively.
Of cours
> On 26 Nov 2017, at 10:00, Jonathan Morton wrote:
>
> On a purely theoretical basis, it probably isn't as fast, because now the
> 'found' condition has to be tested multiple times on the fast and slow paths
> alike. Only if the compiler is smart enough to transform it back the way it
> was
On a purely theoretical basis, it probably isn't as fast, because now the
'found' condition has to be tested multiple times on the fast and slow
paths alike. Only if the compiler is smart enough to transform it back the
way it was...
...and IMHO the goto version is easier to read too.
- Jonathan
Hi All,
If anyone is brave I’ve just done a commit with the src/dst host allocate stuff
minus the gotos. Not even compile tested! Woke up with it in my head. I’m not
convinced it’s really cleaner, absolutely no idea if it’s faster. But it’s
here with all sorts of major caveats :-)
https://g
> On Nov 26, 2017, at 9:42 AM, Dave Taht wrote:
>
> That is a really, really long, and extremely pleasant, way of saying:
>
> "OK, it doesn't crash".
>
> :)
>
> can flenter work with the veth stuff and namespaces?
Maybe a little more than that, but not much more. :) I put it on git for
refe
That is a really, really long, and extremely pleasant, way of saying:
"OK, it doesn't crash".
:)
can flenter work with the veth stuff and namespaces?
On Sun, Nov 26, 2017 at 12:12 AM, Pete Heist wrote:
> I have a script (called flenter) which can run flent with different
> parameter varia
> On Nov 26, 2017, at 6:46 AM, Dave Taht wrote:
>
> Whatever your primary use case is? My biggest concern is that it
> simply not crash - 300 second long tests, 1200 seconds, all night
> long over and over, again, pounding it flat.
>
> rrul_be, rrul, 100 flows, 1000 flows, multiple stations, u
I have a script (called flenter) which can run flent with different parameter
variations and produce an html report. I’m very sorry again that this doesn’t
yet use flent’s batch feature- it was started before I knew about that.
I’ll use it to produce some “rounds” of cake tests, with notes/analy
25 matches
Mail list logo