| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <4A5BC2B6.9020709@itcare.pl>
Date: Tue, 14 Jul 2009 01:26:46 +0200
From: Paweł Staszewski <pstaszewski@...are.pl>
To: Jarek Poplawski <jarkao2@...il.com>
CC: Eric Dumazet <dada1@...mosbay.com>,
Eric Dumazet <eric.dumazet@...il.com>,
Linux Network Development list <netdev@...r.kernel.org>
Subject: Re: weird problem
Jarek Poplawski pisze:
> On Fri, Jul 10, 2009 at 04:47:54PM +0200, Jarek Poplawski wrote:
>
>> On Fri, Jul 10, 2009 at 01:59:00AM +0200, Paweł Staszewski wrote:
>>
>>> Today i make other tests with change of
>>> /proc/sys/net/ipv4/rt_cache_rebuild_count and kernel 2.6.30.1
>>>
>>> And when rt_cache_rebuild_count is set to "-1" i have always load on
>>> x86_64 machine approx 40-50% of each cpu where network card is binded by
>>> irq_aff
>>>
>>> when rt_cache_rebuild_count is set to more than "-1" i have 15 to 20 sec
>>> of 1 to 3% cpu and after 40-50% cpu
>>>
>> ...
>>
>> Here is one more patch for testing (with caution!). It adds possibility
>> to turn off cache disabling (so it should even more resemble 2.6.28)
>> after setting: rt_cache_rebuild_count = 0
>>
>> I'd like you to try this patch:
>> 1) together with the previous patch and "rt_cache_rebuild_count = 0"
>> to check if there is still the difference wrt. 2.6.28; Btw., let
>> me know which /proc/sys/net/ipv4/route/* settings do you need to
>> change and why
>>
>> 2) alone (without the previous patch) and "rt_cache_rebuild_count = 0"
>>
>> 3) if it's possible to try 2.6.30.1 without these patches, but with
>> default /proc/sys/net/ipv4/route/* settings, and higher
>> rt_cache_rebuild_count, e.g. 100; I'm interested if/how long it
>> takes to trigger higher cpu load and the warning "... rebuilds is
>> over limit, route caching disabled"; (Btw., I wonder why you didn't
>> mention about these or maybe also other route caching warnings?)
>>
>
> Here is take 2 to respect setting "rt_cache_rebuild_count = 0" even
> after cache rebuild counter has been increased earlier. (Btw, don't
> forget about this setting after going back to vanilla kernel.)
>
>
Applied to 2.6.30.1
1) With
rt_cache_rebuild_count = 0
grep . /proc/sys/net/ipv4/route/*
/proc/sys/net/ipv4/route/error_burst:1250
/proc/sys/net/ipv4/route/error_cost:250
/proc/sys/net/ipv4/route/gc_elasticity:4
/proc/sys/net/ipv4/route/gc_interval:15
/proc/sys/net/ipv4/route/gc_min_interval:0
/proc/sys/net/ipv4/route/gc_min_interval_ms:0
/proc/sys/net/ipv4/route/gc_thresh:190536
/proc/sys/net/ipv4/route/gc_timeout:15
/proc/sys/net/ipv4/route/max_size:1524288
/proc/sys/net/ipv4/route/min_adv_mss:256
/proc/sys/net/ipv4/route/min_pmtu:552
/proc/sys/net/ipv4/route/mtu_expires:600
/proc/sys/net/ipv4/route/redirect_load:5
/proc/sys/net/ipv4/route/redirect_number:9
/proc/sys/net/ipv4/route/redirect_silence:5120
/proc/sys/net/ipv4/route/secret_interval:3600
I tune this route parameters after looking of traffic/route cache to have not many entries in cache that are not needed anymore
so gc_timeout = 15
limit of max entries = 1524288
And make route cahce a little more "faster" for me after tune
gc_elasticity
secret_interval
gc_interval
gc_thresh
So with this parameters
15 sec of
something like this:
00:41:23 CPU %usr %nice %sys %iowait %irq %soft %steal %guest %idle
00:41:24 all 0.00 0.00 0.12 0.00 1.49 10.46 0.00 0.00 87.92
00:41:24 0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:41:24 1 0.00 0.00 0.00 0.00 4.00 36.00 0.00 0.00 60.00
00:41:24 2 0.00 0.00 0.00 0.00 8.91 47.52 0.00 0.00 43.56
00:41:24 3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:41:24 4 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:41:24 5 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:41:24 6 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:41:24 7 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
and 15 sec of something like this:
00:41:44 CPU %usr %nice %sys %iowait %irq %soft %steal %guest %idle
00:41:45 all 0.00 0.00 0.00 0.00 0.00 0.42 0.00 0.00 99.58
00:41:45 0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:41:45 1 0.00 0.00 0.00 0.00 0.00 1.00 0.00 0.00 99.00
00:41:45 2 0.00 0.00 0.00 0.00 0.00 2.04 0.00 0.00 97.96
00:41:45 3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:41:45 4 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:41:45 5 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:41:45 6 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:41:45 7 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
So i change
/proc/sys/net/ipv4/route/gc_timeout to 1
with
rt_cache_rebuild_count = 0
And output is like 20 sec of something like this
00:48:52 CPU %usr %nice %sys %iowait %irq %soft %steal %guest %idle
00:48:53 all 0.00 0.00 0.19 0.00 0.19 0.58 0.00 0.00 99.03
00:48:53 0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:48:53 1 0.00 0.00 0.99 0.00 0.99 0.00 0.00 0.00 98.02
00:48:53 2 0.00 0.00 0.00 0.00 0.00 2.00 0.00 0.00 98.00
00:48:53 3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:48:53 4 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:48:53 5 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:48:53 6 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:48:53 7 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
and after this two second of something like this:
00:48:49 CPU %usr %nice %sys %iowait %irq %soft %steal %guest %idle
00:48:50 all 0.00 0.00 0.09 0.00 0.27 2.17 0.00 0.00 97.46
00:48:50 0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:48:50 1 0.00 0.00 0.00 0.00 1.96 6.86 0.00 0.00 91.18
00:48:50 2 0.00 0.00 0.00 0.00 0.99 16.83 0.00 0.00 82.18
00:48:50 3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:48:50 4 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:48:50 5 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:48:50 6 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:48:50 7 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:48:50 CPU %usr %nice %sys %iowait %irq %soft %steal %guest %idle
00:48:51 all 0.00 0.00 0.00 0.00 1.86 10.41 0.00 0.00 87.73
00:48:51 0 0.00 0.00 0.00 0.00 0.00 1.00 0.00 0.00 99.00
00:48:51 1 0.00 0.00 0.00 0.00 4.85 26.21 0.00 0.00 68.93
00:48:51 2 0.00 0.00 1.00 0.00 5.00 29.00 0.00 0.00 65.00
00:48:51 3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:48:51 4 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:48:51 5 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:48:51 6 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:48:51 7 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
Another test:
gc_timeout = 1
rt_cache_rebuild_count = 100
10 to 14 sec of something like this:
00:51:36 CPU %usr %nice %sys %iowait %irq %soft %steal %guest %idle
00:51:37 all 0.00 0.00 0.00 0.00 0.00 0.27 0.00 0.00 99.73
00:51:37 0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:51:37 1 0.00 0.00 0.00 0.00 0.00 2.00 0.00 0.00 98.00
00:51:37 2 0.00 0.00 0.00 0.00 0.00 1.00 0.00 0.00 99.00
00:51:37 3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:51:37 4 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:51:37 5 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:51:37 6 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:51:37 7 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
and two seconds of 10 to 30% cpu load more
2).
Only last patch
and almost all the time output like this
00:59:49 CPU %usr %nice %sys %iowait %irq %soft %steal %guest %idle
00:59:50 all 0.00 0.00 0.13 0.00 1.73 8.00 0.00 0.00 90.13
00:59:50 0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:59:50 1 0.00 0.00 0.00 0.00 4.00 24.00 0.00 0.00 72.00
00:59:50 2 0.00 0.00 0.00 0.00 8.91 34.65 0.00 0.00 56.44
00:59:50 3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:59:50 4 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:59:50 5 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:59:50 6 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
00:59:50 7 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
sometimes after 15 to 30 sec i have 1 to 2% cpu load
3).
with default settings and without this patch i have almost all the time output like this:
01:21:40 CPU %usr %nice %sys %iowait %irq %soft %steal %guest %idle
01:21:41 all 0.00 0.00 0.00 0.00 2.14 10.97 0.00 0.00 86.89
01:21:41 0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
01:21:41 1 0.00 0.00 0.00 0.00 6.93 34.65 0.00 0.00 58.42
01:21:41 2 0.00 0.00 0.00 0.00 7.07 42.42 0.00 0.00 50.51
01:21:41 3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
01:21:41 4 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
01:21:41 5 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
01:21:41 6 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
01:21:41 7 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
with my settings:
/proc/sys/net/ipv4/route/error_burst:1250
/proc/sys/net/ipv4/route/error_cost:250
/proc/sys/net/ipv4/route/gc_elasticity:4
/proc/sys/net/ipv4/route/gc_interval:15
/proc/sys/net/ipv4/route/gc_min_interval:0
/proc/sys/net/ipv4/route/gc_min_interval_ms:0
/proc/sys/net/ipv4/route/gc_thresh:190536
/proc/sys/net/ipv4/route/gc_timeout:15
/proc/sys/net/ipv4/route/max_size:1524288
/proc/sys/net/ipv4/route/min_adv_mss:256
/proc/sys/net/ipv4/route/min_pmtu:552
/proc/sys/net/ipv4/route/mtu_expires:600
/proc/sys/net/ipv4/route/redirect_load:5
/proc/sys/net/ipv4/route/redirect_number:9
/proc/sys/net/ipv4/route/redirect_silence:5120
/proc/sys/net/ipv4/route/secret_interval:3600
15 sec of 30 to 50 % cpu and 15 sec 1 to 2 % cpu
with
/proc/sys/net/ipv4/route/gc_interval:1
almost all the time like this
01:23:45 CPU %usr %nice %sys %iowait %irq %soft %steal %guest %idle
01:23:46 all 0.00 0.00 0.00 0.00 0.00 0.12 0.00 0.00 99.88
01:23:46 0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
01:23:46 1 0.00 0.00 0.00 0.00 1.00 0.00 0.00 0.00 99.00
01:23:46 2 0.00 0.00 0.00 0.00 0.00 1.02 0.00 0.00 98.98
01:23:46 3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
01:23:46 4 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
01:23:46 5 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
01:23:46 6 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
01:23:46 7 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
with max two outputs of 20 to 30% cpu in different times from 12 to 15sec
And i dont know but i think patch for turning off route cache is not working
because with this patches and
rt_cache_rebuild_count = 0
I have
rtstat -k entries -c 10 -i 1
rt_cache|
entries|
52929|
46301|
40971|
36928|
34882|
34915|
36431|
37752|
Regards
Pawel Staszewski
> Jarek P.
> --- (debugging patch #2 take 2; apply to 2.6.30.1 or 2.6.29.6)
>
>
> net/ipv4/route.c | 21 ++++++++++++++-------
> 1 files changed, 14 insertions(+), 7 deletions(-)
>
> diff --git a/net/ipv4/route.c b/net/ipv4/route.c
> index 278f46f..f74db20 100644
> --- a/net/ipv4/route.c
> +++ b/net/ipv4/route.c
> @@ -678,8 +678,9 @@ static inline u32 rt_score(struct rtable *rt)
>
> static inline bool rt_caching(const struct net *net)
> {
> - return net->ipv4.current_rt_cache_rebuild_count <=
> - net->ipv4.sysctl_rt_cache_rebuild_count;
> + return (net->ipv4.current_rt_cache_rebuild_count <=
> + net->ipv4.sysctl_rt_cache_rebuild_count) ||
> + net->ipv4.sysctl_rt_cache_rebuild_count == 0;
> }
>
> static inline bool compare_hash_inputs(const struct flowi *fl1,
> @@ -1181,12 +1182,18 @@ restart:
> } else {
> if (chain_length > rt_chain_length_max) {
> struct net *net = dev_net(rt->u.dst.dev);
> - int num = ++net->ipv4.current_rt_cache_rebuild_count;
> - if (!rt_caching(dev_net(rt->u.dst.dev))) {
> - printk(KERN_WARNING "%s: %d rebuilds is over limit, route caching disabled\n",
> - rt->u.dst.dev->name, num);
> +
> + if (net->ipv4.sysctl_rt_cache_rebuild_count > 0) {
> + int num = ++net->ipv4.current_rt_cache_rebuild_count;
> +
> + if (!rt_caching(net))
> + printk(KERN_WARNING
> + "%s: %d rebuilds is over limit, "
> + "route caching disabled\n",
> + rt->u.dst.dev->name, num);
> +
> + rt_emergency_hash_rebuild(net);
> }
> - rt_emergency_hash_rebuild(dev_net(rt->u.dst.dev));
> }
> }
>
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@...r.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
>
>
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists