Search String: Display: Description: Sort:

Results:

References: [ +subject:/^(?:^\s*(re|sv|fwd|fw)[\[\]\d]*[:>-]+\s*)*\[Lse\-tech\]\s+Re\:\s+A\s+common\s+layer\s+for\s+Accounting\s+packages\s*$/: 52 ]

Total 52 documents matching your query.

1. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: Guillaume Thouvenin <guillaume.thouvenin@xxxxxxxx>
Date: Tue, 01 Mar 2005 09:21:39 +0100
I tested without user space listeners and the cost is negligible. I will test with a user space listeners and see the results. I'm going to run the test this week after improving the mechanism that s
/archives/netdev/2005-03/msg00000.html (11,881 bytes)

2. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: Kaigai Kohei <kaigai@xxxxxxxxxxxxx>
Date: Tue, 01 Mar 2005 22:38:25 +0900
I'm also trying to mesure the process-creation/destruction performance on following three environment. Archtechture: i686 / Distribution: Fedora Core 3 * Kernel Preemption is DISABLE * SMP kernel bu
/archives/netdev/2005-03/msg00004.html (12,162 bytes)

3. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: Guillaume Thouvenin <guillaume.thouvenin@xxxxxxxx>
Date: Tue, 01 Mar 2005 14:53:56 +0100
I don't see this limit on my computer. I'm currently running the lmbench with a new fork connector patch (one that enable/disable fork connector) on an SMP computer. I will send results and the new p
/archives/netdev/2005-03/msg00007.html (13,430 bytes)

4. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: Evgeniy Polyakov <johnpol@xxxxxxxxxxx>
Date: Tue, 01 Mar 2005 17:17:57 +0300
2.6.11-rc4-mm1 tree does not have the latest connector. Various fixes were added, not only that. I run the latest patch Guillaume sent to me(with small updates), fork bomb with more than 100k forks p
/archives/netdev/2005-03/msg00009.html (15,069 bytes)

5. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: roen@xxxxxxxxx>
Date: Tue, 1 Mar 2005 12:40:41 -0800
I think the hope was to reduce the cost of the accounting hook in fork to "next-to-zero" if accounting is not being used on that system. See Andrew's query earlier: Presumably sending an ignored pack
/archives/netdev/2005-03/msg00028.html (11,064 bytes)

6. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: xxxxxxxxxxxxxx>
Date: Tue, 1 Mar 2005 20:50:33 -0800
Just a thought - perhaps you could see if Jay can test the performance scaling of these changes on larger systems (8 to 64 CPUs, give or take, small for SGI, but big for some vendors.) Things like a
/archives/netdev/2005-03/msg00036.html (12,035 bytes)

7. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: venin@xxxxxxxx>
Date: Wed, 02 Mar 2005 09:58:13 +0100
So I ran the lmbench with three different kernels with the fork connector patch I just sent. Results are attached at the end of the mail and there are three different lines which are: o First line is
/archives/netdev/2005-03/msg00053.html (18,570 bytes)

8. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: venin@xxxxxxxx>
Date: Wed, 2 Mar 2005 01:06:14 -0800
This is the interesting bit, yes? 5-10% slowdown on fork is expected, but why was exec slower? What does "The user space application listened during 15h" mean?
/archives/netdev/2005-03/msg00054.html (12,517 bytes)

9. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: <akpm@xxxxxxxx>
Date: Wed, 02 Mar 2005 10:25:49 +0100
I can't explain it for the moment. I will run test more than once to see if this difference is still here. It means that I ran the user space application before the test and stop it 15 hours later (t
/archives/netdev/2005-03/msg00055.html (18,704 bytes)

10. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: roen@xxxxxxxxx>
Date: Wed, 2 Mar 2005 07:30:45 -0800
Thanks for the summary, Andrew. Guillaume (or anyone else tempted to do this) - it's a good idea, when posting 100 lines of data, to summarize with a line or two of words, as Andrew did here. It is f
/archives/netdev/2005-03/msg00070.html (11,703 bytes)

11. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: xxx>
Date: Mon, 28 Feb 2005 10:59:06 +0900
Does it mean "netlink packets shouled be sent to userspace unconditionally." ? I have advocated steadfastly that fork/exec/exit hooks is necessary to support process-grouping and to account per proce
/archives/netdev/2005-02/msg00958.html (13,911 bytes)

12. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: xxxxxxxxxxxxxxxxxx>
Date: Mon, 28 Feb 2005 03:32:06 +0100
First of all, I'm not aware of the whole discussion, ignore this if it has been brought to attention already. The easiest way is to use netlink_broadcast() and have userspace register to a netlink mu
/archives/netdev/2005-02/msg00960.html (11,024 bytes)

13. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: igai Kohei <kaigai@xxxxxxxxxxxxx>
Date: Mon, 28 Feb 2005 08:17:31 +0300
Kernel accounting already was discussed in lkml week ago - I'm quite sure Guillaume Thouvenin created exactly that. His module creates do_fork() hook and broadcasts various process' states over netli
/archives/netdev/2005-02/msg00965.html (11,979 bytes)

14. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: llmann@xxxxxxxx>
Date: Mon, 28 Feb 2005 08:20:36 +0100
I wrote a new fork connector patch with a callback to enable/disable messages in case there is or isn't listener. I will post it this week. Basically there is a global variable that is manipulated wi
/archives/netdev/2005-02/msg00969.html (12,616 bytes)

15. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: x>
Date: Sun, 27 Feb 2005 23:39:43 -0800
Yes. But this problem can be solved in userspace, with a little library function and a bit of locking. IOW: use the library to enable/disable the fork connector rather than directly doing syscalls. I
/archives/netdev/2005-02/msg00970.html (11,798 bytes)

16. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: xxx>
Date: Mon, 28 Feb 2005 11:04:36 +0300
Why dont just extend protocol a bit? Add header after cn_msg, which will have get/set field and that is all. Properly using seq/ack fields userspace can avoid locks. -- Evgeniy Polyakov Crash is bett
/archives/netdev/2005-02/msg00974.html (13,236 bytes)

17. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: awal@xxxxxxxxx>
Date: 28 Feb 2005 07:10:58 -0500
Havent seen the beginnings of this thread. But whatever you are trying to do seems to suggest some complexity that you are trying to workaround. What was wrong with just going ahead and just always i
/archives/netdev/2005-02/msg00979.html (13,760 bytes)

18. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: alicchio <ecalicchio@xxxxxxxx>
Date: Mon, 28 Feb 2005 14:20:51 +0100
I guess parts of the wheel are broken and need to be reinvented ;-> Additional you may want to extend netlink a bit to check whether there is a listener before creating the messages. The method to d
/archives/netdev/2005-02/msg00980.html (12,297 bytes)

19. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: x>
Date: 28 Feb 2005 08:40:10 -0500
netlink broadcast or a wrapper around it. Why even bother doing the check with netlink_has_listeners()? cheers, jamal
/archives/netdev/2005-02/msg00981.html (13,295 bytes)

20. Re: [Lse-tech] Re: A common layer for Accounting packages (score: 1)
Author: npol@xxxxxxxxxxx>
Date: Mon, 28 Feb 2005 14:53:07 +0100
* jamal <1109598010.2188.994.camel@xxxxxxxxxxxxxxxx> 2005-02-28 08:40 To implement the master enable/disable switch they want. The messages don't get send out anyway but why bother doing all the work
/archives/netdev/2005-02/msg00982.html (11,582 bytes)


This search system is powered by Namazu