Mail Archive Search: <+subject:/^(?:^\s*(re|sv|fwd|fw)[\[\]\d]*[:>-]+\s*)*route\s+cache\s+DoS\s+testing\s+and\s+softirqs\s*$/>

Results:

References: [ +subject:/^(?:^\s*(re|sv|fwd|fw)[\[\]\d]*[:>-]+\s*)*route\s+cache\s+DoS\s+testing\s+and\s+softirqs\s*$/: 88 ]

Total 88 documents matching your query.

1. Re: route cache DoS testing and softirqs (score: 1): Author: Andrea Arcangeli <andrea@xxxxxxx>; Date: Thu, 1 Apr 2004 15:16:57 +0200; This sounds reasonable. However as a start I was thinking at having hardirq run only the softirq they posted actively, and local_bh_enable run only the softirq that have been posted by the critical s; /archives/netdev/2004-04/msg00003.html (9,950 bytes)
2. Re: route cache DoS testing and softirqs (score: 1): Author: Robert Olsson <Robert.Olsson@xxxxxxxxxxx>; Date: Thu, 1 Apr 2004 15:44:57 +0200; How would softirq's scheduled be softirq's be run? For a definitive solution a think Alexey analysis make sense: "All of them happening outside of ksoftirqd are equally bad, unaccountable, uncontroll; /archives/netdev/2004-04/msg00004.html (12,038 bytes)
3. Re: route cache DoS testing and softirqs (score: 1): Author: Robert Olsson <Robert.Olsson@xxxxxxxxxxx>; Date: Wed, 31 Mar 2004 00:33:11 +0200; Well we have to ask Dipankar... But I'll buy a beer if it's not on. :) Anyway w. NAPI enabled. 2 * 304 kpps DoS flows into eth0, eth2. Flows are 2 * 10 Millions 64 byte pkts. 32 k buckets routehash.; /archives/netdev/2004-04/msg00008.html (10,875 bytes)
4. Re: route cache DoS testing and softirqs (score: 1): Author: Robert Olsson <Robert.Olsson@xxxxxxxxxxx>; Date: Mon, 5 Apr 2004 19:11:52 +0200; Hello! Yes it seems reduce RCU latency in our setup as well. It does not eliminate overflows but reduces with ~50% and increases the throughput a bit. dst cache overflow depends on RCU-delay + gc_min; /archives/netdev/2004-04/msg00054.html (12,298 bytes)
5. Re: route cache DoS testing and softirqs (score: 1): Author: Dipankar Sarma <dipankar@xxxxxxxxxx>; Date: Tue, 6 Apr 2004 02:52:20 +0530; Looks better atleast. Can you apply the following patch (rs-throttle-rcu) on top of rcu-softirq.patch in your tree and see if helps a little bit more ? Please make sure to set the kernel paramenters; /archives/netdev/2004-04/msg00060.html (19,779 bytes)
6. Re: route cache DoS testing and softirqs (score: 1): Author: Robert Olsson <Robert.Olsson@xxxxxxxxxxx>; Date: Tue, 6 Apr 2004 14:55:19 +0200; OK! Well not tested yet but I don't think we will get rid overflow totally in my setup. I've done a little experimental patch so *all* softirq's are run via ksoftirqd. total droppped tsquz throttl bh; /archives/netdev/2004-04/msg00085.html (11,331 bytes)
7. Re: route cache DoS testing and softirqs (score: 1): Author: Dipankar Sarma <dipankar@xxxxxxxxxx>; Date: Wed, 7 Apr 2004 01:22:49 +0530; Robert, you should try out rs-throttle-rcu.patch. The idea is that we don't run too many callbacks in a single rcu. In my setup, at 100kpps, I see as many as 30000 rcu callbacks in a single tasklet h; /archives/netdev/2004-04/msg00095.html (11,275 bytes)
8. Re: route cache DoS testing and softirqs (score: 1): Author: Robert Olsson <Robert.Olsson@xxxxxxxxxxx>; Date: Wed, 7 Apr 2004 17:23:28 +0200; Hello! No bad things happens, lots of overflows and drop in performance and the userland app can stall for 32 sec. We seems to spin in softirq to much and still don't get things done. Cheers. --ro Wi; /archives/netdev/2004-04/msg00114.html (12,317 bytes)
9. Re: route cache DoS testing and softirqs (score: 1): Author: Dipankar Sarma <dipankar@xxxxxxxxxx>; Date: Thu, 8 Apr 2004 01:18:43 +0530; Argh!! Andrea, this means that throttling rcu callbacks with back-to-back rcu tasklets for better scheduling latency is bad for this kind of DoS situation. I think we will have to address the softirq; /archives/netdev/2004-04/msg00119.html (12,322 bytes)
10. Re: route cache DoS testing and softirqs (score: 1): Author: kuznet@xxxxxxxxxxxxx; Date: Thu, 8 Apr 2004 17:29:13 +0400 (MSD); BTW what's about performance in this extremal situation? Also, Robert, let's count the numbers again. With this change you should have latency much less 100msec when priority of ksoftirqd is high. S; /archives/netdev/2004-04/msg00127.html (9,328 bytes)
11. Re: route cache DoS testing and softirqs (score: 1): Author: kuznet@xxxxxxxxxxxxx; Date: Thu, 8 Apr 2004 17:38:27 +0400 (MSD); To all that I remember Ingo insisted that the run from local_bh_enable() is necessary for softirqs triggered by hardirqs while softirqs are disabled. Otherwise he observed random delays and redundan; /archives/netdev/2004-04/msg00128.html (9,212 bytes)
12. Re: route cache DoS testing and softirqs (score: 1): Author: Robert Olsson <Robert.Olsson@xxxxxxxxxxx>; Date: Thu, 8 Apr 2004 16:07:31 +0200; First I used the patch to defer all softirq's to ksoftirq with call_rcu_bh() patch. Sofar this has been the best combination giving both pure sofirq performance and also good response from the userla; /archives/netdev/2004-04/msg00129.html (10,298 bytes)
13. route cache DoS testing and softirqs (score: 1): Author: xxxxxxxxx>; Date: Tue, 30 Mar 2004 00:15:50 +0530; Robert Olsson noticed dst cache overflows while doing DoS stress testing in a 2.6 based router setup a few months and davem, alexey, robert and I have been discussing this privately since then (198 m; /archives/netdev/2004-03/msg00567.html (12,603 bytes)
14. Re: route cache DoS testing and softirqs (score: 1): Author: da@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>; Date: Tue, 30 Mar 2004 00:29:26 +0200; softirqs are already capable of being offloaded to scheduler-friendy kernel threads to avoid starvation, if this wasn't the case NAPI would have no way to work in the first place and everything else; /archives/netdev/2004-03/msg00576.html (19,692 bytes)
15. Re: route cache DoS testing and softirqs (score: 1): Author: "David S. Miller" <davem@xxxxxxxxxx>; Date: Tue, 30 Mar 2004 10:36:14 +0530; I think set_current_state(TASK_INTERRUPTIBLE) before kthread_should_stop() _is_ required, otherwise kthread_stop can fail to destroy a kthread. kthread_stop does: 1. kthread_stop_info.k = k; 2. wake_; /archives/netdev/2004-03/msg00583.html (11,157 bytes)
16. Re: route cache DoS testing and softirqs (score: 1): Author: giri <vatsa@xxxxxxxxxx>; Date: Tue, 30 Mar 2004 11:05:15 +0530; hmm .. I meant a) happening _before_ 1) and b) happening _after_ 2) .. Again I meant "even if b) happens _before_ 1) and c) _after_ 2) !! -- Thanks and Regards, Srivatsa Vaddagiri, Linux Technology C; /archives/netdev/2004-03/msg00585.html (11,278 bytes)
17. Re: route cache DoS testing and softirqs (score: 1): Author: omar <nagendra_tomar@xxxxxxxxxxx>; Date: Tue, 30 Mar 2004 20:13:24 +0530; Hmm.. What about firewalls and routers on the internet ? Shouldn't they care ? Tried it and it didn't work. I still got dst cache overflows. I will dig out more numbers about what what happened - is; /archives/netdev/2004-03/msg00589.html (12,134 bytes)
18. Re: route cache DoS testing and softirqs (score: 1): Author: uf Goolamabbas <yusufg@xxxxxxxxxxxx>; Date: Tue, 30 Mar 2004 17:11:48 +0200; you're right, I had a private email exchange with Andrew about this yesterday, he promptly pointed it out too. But my point is that the previous code was broken too, it wasn't me breaking the code, I; /archives/netdev/2004-03/msg00590.html (12,118 bytes)
19. Re: route cache DoS testing and softirqs (score: 1): Author: x>; Date: Wed, 31 Mar 2004 01:23:15 +0530; It doesn't look as if we are processing much from ksoftirqd at all in this case. I did the following instrumentation - if (in_interrupt() && local_softirqd_running()) return; max_restart = MAX_SOFTIR; /archives/netdev/2004-03/msg00591.html (13,373 bytes)
20. Re: route cache DoS testing and softirqs (score: 1): Author: x>; Date: Wed, 31 Mar 2004 00:05:05 +0400 (MSD); There are no hardirqs in the case under investigation, remember? What's about the problem it really splits to two ones: 1. The _new_ problem when bad latency of rcu hurts core functionality. This is; /archives/netdev/2004-03/msg00592.html (13,263 bytes)

Current List: 1 - 20
Page: [1] [2] [3] [4] [5]