Re: Compatible NIC
In a message dated 11/1/04 1:36:24 AM Eastern Standard Time, [EMAIL PROTECTED] writes: On Sunday 31 October 2004 08:54 am, [EMAIL PROTECTED] wrote: I know who Kris is. I respect and appreciate his contributions. I don't respect being lied to. And I don't respect the unconditional rejection of criticism by the team. This is not the appropriate venue for that sort of conversation. Please take it elsewhere Yes Use 5.x! is not technical help. Don't use 5.x because its slow IS technical help. You guys just dont want anyone to say it. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Compatible NIC
In a message dated 11/1/04 12:47:19 PM Eastern Standard Time, [EMAIL PROTECTED] writes: Yes Use 5.x! is not technical help. Don't use 5.x because its slow IS technical help. You guys just dont want anyone to say it. I'm not about to get into whether or not these 2 things are or are not technical help. But I certainly don't think that anyone does want to say that saying Don't use 5.x because its slow is technical help because many of us see it as largely unfounded opinion. I have seen lots of posts trying to tell you that 5.x is just as fast or faster and demanding benchmarks. While I note that you have not provided any grounds on which to make the claim of decreased speed on similarly configured software running on matching (or better, the same) hardware. --- Why don't you want to get into it, since its your entire point? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Compatible NIC
In a message dated 10/31/04 5:00:35 AM Eastern Standard Time, [EMAIL PROTECTED] writes: No one is forcing you to read anything. I never copy you on anythere, I'm not sure what anythere means, but by sending messages to the mailing list, you're copying Kris and everybody else who is on the list. You don't actually read every message on every list you are on, do you? If you do then you really need a hobby. I scan the subjects and read the 2 or 3 out of 50 that sound interesting. I suggest you do the same. A member of the Gustapo said: Kris was just asking you to respect the charter of the mailing list. I've looked at about 15 messages from you, most of them insulting, some voicing opinions that run contrary to fact, and suggesting that you have a good overall understanding of the project. Given that you don't know who Kris is, it's difficult to believe the last point. --- So the charter of the mailing list is that only good and positive things can be said about FreeBSD, and no one is allowed to make distinctions between good and bad code and/or drivers? Is the soviet union back or what? The charter of this list is for people who want answers about FreeBSD to be able to get them. I felt it necessary to join when I noticed that EVERYONE on the list cheerfully steers poor suckers into using 5.x, even though it appears, after having to beat it out of them, everyone pretty much admits that 5.x isn't better than 4.x at the moment, and that even 5.3 is going to be a lower performer. To me, spinning the tale of 5.x to those in search of real answers is violating the charter, unless the charter has changed to shamelessly steering everyone to use 5.x for internal, political purposes. I know who Kris is. I respect and appreciate his contributions. I don't respect being lied to. And I don't respect the unconditional rejection of criticism by the team. And I haven't seen any evidence that anyone really has a clue as to how to measure the performance of the product they're developing. I was ridiculed for tearing apart the only test results posted, yet no credible ones were offered. I was asked for an explanation as to why I question drivers written by a certain developer, and I provided the info. Instead of credible counterpoint, I was told that I was wasting people's time. How am I wasting someone's time when I'm telling them not to use drivers that very likely have flaws? How are you helping someone by cheerfully recommending things known to be poorly done? To not hurt the developer's feelings? Is this forum about helping users or about coddling developers? Optimizing a driver is as important as making it bug free. If they do a half-assed job then criticism is warranted. Its easy to dismiss people who ask hard questions as trolls. Its a lot more difficult to answer the questions credibly. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: mysql and system/nice cpu usage
In a message dated 10/31/04 11:03:12 AM Eastern Standard Time, [EMAIL PROTECTED] writes: but never come across with a problem like that. I am thinking to use Zend Optimizer. Maybe that helps me .. If that doesn't help I was thinking to run sql on a seperate machine. You might try tuning kern.vm.kmem.size if thats not in the tuning suggestions. The OS tends to allocate way more memory than needed for the kernel if you have a lot of memory in the system; you probably don't need more than 100M or so unless you're running bgp or something unusual. Once you start swapping with mysql and php you're dead. Moving to another system can help, but be aware that if your network is busy it can add some different inefficiencies. If you do go to a separate system connect it with a dedicated NIC if possible, to alleviate network backup. If you have a multiple bus machine, moving your NIC to a separate bus from the HDD can significantly increase performance. When you have the NIC and HDD on the same bus, heavy network traffic can cause disk operations to back up and substantially slow database applications. Make sure the busses are really separate (and not cascaded), otherwise it won't help. Also if you're on a 32bit bus machine you'll have a lot more contention than with a pci-x bus. Most people think that if you have enough bus then it doesnt matter, but thats dead wrong. Bus contention between devices is a major performance factor. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: 5.3-RC1 - Hangs on high net load(?)
In a message dated 10/31/04 11:56:29 AM Eastern Standard Time, [EMAIL PROTECTED] writes: After upgrading from 5.2.1 to 5.3-RC1 (including a recompile of all ports), I noticed my system locking up when there was a high network load. A common way to reproduce this is to run a BiTTorrent client and, say, copy a file to another local machine over NFS. The computer does not respond to pings or any key presses. I'm using the lnc(4) driver for an AMD PCnet PCI ethernet card (exact model escapes me) [Am79C970/1/2/3/5/6 PCnet LANCE PCI Ethernet Controller]. (I'm aware of the pcn(4) driver, but cannot get it to recognize the card). I can reproduce this with both the 4BSD and ULE schedulers, if that matters. All relevant sysctl variables are defaults. Is this a known issue? Any ideas? Thanks. Are you certain that its not NFS that's locking up? Certainly the use of NFS muddies the issue, as it doesnt like losing packets and isn't very eloquent in its handling of adversity. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: mysql and system/nice cpu usage
In a message dated 10/31/04 12:35:23 PM Eastern Standard Time, [EMAIL PROTECTED] writes: Hi The oid you are talking about is not valid in FreeBSD-4. Maybe you are talking about FreeBSD-5 sysctl oids? But it does worth to try but I am not sure which oid it is in FreeBSD 4.. REGARDS PS: I have found vm.kvm_size. I think it is the one that corresponds in your email? - No, I gave you the correct one. Why dont you do some googling rather than trying to sift through docs? Its a kernel-level oid. Look in /boot/defaults/loader.conf As for the memory output, perhaps one of the geniuses that came up with that cryptic output can help ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: 5.3-RC1 - Hangs on high net load(?)
Are you certain that its not NFS that's locking up? Certainly the use of NFS muddies the issue, as it doesnt like losing packets and isn't very eloquent in its handling of adversity. I'm positive. The machine itself locks up. I cannot ctrl+alt+del, can't switch VCs; it freezes. NFS isn't necessarily a part of the problem anyway. Just general net load seems to do it. For example, if I start multiple downloads with using Bittorent. I haven't narrowed the issue down terribly far, though. It did not hang while grabbing an ISO via HTTP at ~300 KBps. It does, however, hang when there's significant load and Bittorent is part of the formula. But it doesn't _always_ hang when using it. Have you ruled out livelock? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: GPL vs BSD Licence
Again, the reality is that none of this (the existence of some products that exist as binary modules) harm the community. They offer choices for users, and the more choices the better. What a horrible place the world would be without TiVo (who never would have done the work if they couldn't protect it) Yah, people might actually have to READ a book instead of watching the TV. What horrors!! Ted Or watch Meet the Press, and actually HEAR what your leaders say with context, instead of relying on whatever the media wants you to hear, or what fits an author's private agenda. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: 7520 Chipset support in 4.x
Many of the new MBs from such tiny vendors as Dell and Supermicro are based on the 7520, and word is that FreeBSD 4.x doesn't support it. Is support forthcoming? We have 2 Dell PowerEdge 1850 servers which have the e7520 chipset. They hang consistently in 4.10-RELEASE and below whenever there is high network or disk utilization. We have not been able to get any debugging info. After upgrading to 4.10-STABLE a couple of weeks ago, they no longer hang, but they are _really_ slow to perform network and disk operations. They work fine in FreeBSD 5.3, but unfortunately our applications do not run without recompiling. We do not want to change our environment to support different binaries for different machines, and we don't want to use 5.X in production until it is STABLE. I want to echo the above question. Are there patches available or forthcoming to fix the problems with the e7520? Thanks, - Rob Watt I think that we can imply from the lack of response on this subject that 4.x is not really still supported, since just about all of the new motherboards for Intel processors from leaders Dell and Supermicro are based on the 7520. So, ironically, in order to use the newer, faster processors with FreeBSD, you have to use the newer, slower version of the O/S. Yikes! ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: 7520 Chipset support in 4.x
As in your previous post on the subject, I find it no where near as slow as you have stated. For one who couldn't figure out how to compile without the witness options and various other debug stuff into the kernel and base system, it prolly would be slower. After I took this stuff out of the build my benches were greatly improved, but alas, not to 4.10 speeds. The POINT is not how much slower 5.3 is than 4.10 is, its the fact that a MAJOR chipset isnt well supported in what is SUPPOSED to be the mainstream, stable release. I know its difficult for you to stay on point, but at least try to get the point before ranting about the subject. Since you just babble about your tests but have never shown any results, no one really knows what your tests test, if anything. The only test anyone has posted was completely bogus (some nonsense about firing packets through a socket interface), so you just can't say my tests show unless you clarify what you have done. When 5.3 is released we can banter about benchmarks, as it seems pointless to do it now since its not done. It may be pointless anyway, since they've pretty much admitted that 5.3 isn't (yet) going to rival the efficiency of 4.x. I can accept that, but if 5.x isnt ready, then important chipsets should be supported in 4.x BEFORE they are supported in 5, not when someone gets around to backporting it. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: dummynet
In a message dated 10/29/04 8:26:51 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: The boss pays his sysadmin every week, no matter what. The Boss expects that the systems will runs with the least overall cost. Sometimes that means buying something, sometimes that means configuring what is there. Unfortunately most ISPs don't know much about business, so I guess explaining the concept of opportunity costs to you would be a waste of time. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: GPL vs BSD Licence
In a message dated 10/29/04 5:27:20 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: Then you either know nothing about programming or nothing about their products. Do you think they do gigabit bandwidth management, with features not in the kernel, from user space? That's not what I meant and not what I wrote. you can write a loadable kernel module w/o changing the kernel sources, can't you? Not without adding hooks, which which would have to be provided under the GPL. You've obviously never done anything like this, or know how it works, so why do you feel qualitifed to comment on it? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: GPL vs BSD Licence
In a message dated 10/29/04 2:10:23 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: the GPL. I seem to recall the discussion was about nVidia's closed source, binary only drivers but, according to Linus, affects all similar products. I'm unsure if and how this issue is being dealt with. It is. It is the stated policy of the FSF that loadable kernel modules are considered part of the GPL work and therefore must be GPL'ed themselves. That is where all this is coming from. It is kind of a personal vendetta/issue with RMS I understand. This position has also created lots of controversy as you might imagine. The FSF doesnt have standing with Linux so they can blow as hard as they like and no one will really care. The FSF is a bunch of weenies whos only mission in life is to abolish anything thats not open source. Linus has stated that, if software was written for a different O/S and was ported to linux, its not a derivative work and binary modules are acceptable and don't have to be GPLed Again, the reality is that none of this (the existence of some products that exist as binary modules) harm the community. They offer choices for users, and the more choices the better. What a horrible place the world would be without TiVo (who never would have done the work if they couldn't protect it) ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: GPL vs BSD Licence
In a message dated 10/28/04 9:16:02 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: But then, I'm not sure (and I mean it) if there can be any piece of software which, if designed for e.g. Linux, can be written w/o using any system headers, libraries or whatsoever. -- I find it impossible for any reasonable person to believe that, by making its header files available an O/S vendor therefore owns the rights to anything that runs on, or interoperates with the O/S. So Microsoft owns Photoshop. And Netscape too. So why are they fighting? Its been fairly well established that Lawyers know a lot about law but not much about computing. You can't apply copyright law verbatim to an operating system, because unlike a written work, the operating systems entire purpose is to provide hooks for external applications and device drivers. Claiming that anything that works with it is a derivative is, quite simply, ridiculous. The GPL is a myth. It will never be tested because if it is, it will lose all of its teeth. Its much more useful in a speculative state. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: GPL vs BSD Licence
In a message dated 10/29/04 2:12:21 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: Linus is just a big dope anyway, so who cares what he thinks? He's like Kerry. He thinks whatever is convenient for him to think at the time. And RMS is a lot like Bush who says whatever is convenient for him to say at the time. There - now we got the election dragged in! Shall we try for the Nazi's? ;-) Nah, we'll make the Jewish people uncomfortable. Lets talk about the French. Everyone hates the French. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Compatible NIC
In a message dated 10/29/04 12:13:14 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: quick sho I want to buy a NIC and I want it to be compatible with FreeBSD. Is RealTek 8139 compatible with FreeBsd ? rt answer : yes long answer: see hardware notes' it is listed there. RealTek 8129/8139 Fast Ethernet NICs ( rl(4) driver) Check the driver source. Any driver witten by Bill Paul should be avoided if possible. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
(no subject)
The problem with dummynet is that once you do all the work and figure it all out, its still only marginally functional compared to something relatively inexpensive. So instead of buying the $3500 box that is everything you need, you've spend $800 on hardware, $2000 worth of time, and you still have something not nearly as good. One question, have you ever used dummynet? If so, I'm curious as to why you find it only marginal. Not to be rude, but if you've not used it, please stop trolling. -- Micheal Patterson TSG Network Administration 405-917-0600 One can tell by looking at the code that it won't scale. And I know more than 20 people who've been bitten on the butt by trying to use it, and then buying something when they hit the wall with it, or finding out it can't do what they need. The question is, have YOU used anything else? Or are you like the old woman who still washes her clothes in the river because those darned mechanical things aren't worth it? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: dummynet
The problem with dummynet is that once you do all the work and figure it all out, its still only marginally functional compared to something relatively inexpensive. So instead of buying the $3500 box that is everything you need, you've spend $800 on hardware, $2000 worth of time, and you still have something not nearly as good. One question, have you ever used dummynet? If so, I'm curious as to why you find it only marginal. Not to be rude, but if you've not used it, please stop trolling. -- Micheal Patterson TSG Network Administration 405-917-0600 One can tell by looking at the code that it won't scale. And I know more than 20 people who've been bitten on the butt by trying to use it, and then buying something when they hit the wall with it, or finding out it can't do what they need. The question is, have YOU used anything else? Or are you like the old woman who still washes her clothes in the river because those darned mechanical things aren't worth it? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Compatible NIC
In a message dated 10/29/04 1:13:05 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: RealTek 8129/8139 Fast Ethernet NICs ( rl(4) driver) Check the driver source. Any driver witten by Bill Paul should be avoided if possible. What the hell man. Do you have anything better to do than troll? Every responce I have read from you seems something written out of a ad or some buzz of slashdot or something you saw and thought you would spew your techy knowledge. If you have a specific reason you feel a specific driver should be avoided, please voice it. Ok. Bill Paul is an academic who was contracted to write a zillion drivers for FreeBSD. He basically cut and pasted them all, barely tested them, and when he was done he didn't support them. Read his comments and you'll know that he was more concerned with getting done than doing a good job. His goal was just to do them. I've tested most of them. Have you? Have you done anything except listen to a bunch of idiots on this list who are as far from a real network engineers that they can be? Bill Paul himself proclaimed that the realtek chip is the worst controller ever created, but you idiots wholeheartedly recommend it, as if ethernet cards and drivers are generic, interchangeable parts. You guys claim to help people. Part of helping them is steering them away from bad drivers, bad hardware and bad versions of the OS, IMO. I just voiced my opinion. If you want to use them, feel free. Use 5.2.1. with the rl driver. Have the slowest server on the block. What do I care? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: GPL vs BSD Licence
In a message dated 10/29/04 12:38:10 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: The GPL and Linux don't care if you link into their system libraries, they expect that which is why the system libraries are LGPLd ... If I write a piece of code that uses a defined interface, it's utterly preposterous to argue that it is derivative from an *implementation* of that interface, since it could be used with *any* implementation of that interface. Its equally preposterous for the GPLers to claim that anything that works with any O/S is owned by the owner of the OS as a derivative work. But they do, and they will, because it suits them. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: GPL vs BSD Licence
In a message dated 10/29/04 3:54:44 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: Its equally preposterous for the GPLers to claim that anything that works with any O/S is owned by the owner of the OS as a derivative work. But they do, and they will, because it suits them. It is not just the GPL folks. SCO is doing the same thing to IBMs code. Code totally removed from SCOs SysVR4 code is being claimed by SCO as a derivative work. I am not trying to open up a discussion on SCO. Just to point out that this phenomenon is not restricted to the GPL fanatics. The whole SCO vs IBM mess is an illustration of just how goofy the entire GPL is. IBM claims that they have standing to sue over LINUX because they've made contributions. But since contributions are derivative works, then they shouldn't have any copyright. So does everyone who has contributed to linux have standing to sue then? Of course, I think device drivers and modules are different. The entire purpose of the device driver/module interface is to make add-ons possible without having to modify the kernel. The concept that they are also derivative works is nonsense. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Compatible NIC
In a message dated 10/29/04 3:27:34 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: I just voiced my opinion. If you want to use them, feel free. Use 5.2.1. with the rl driver. Have the slowest server on the block. What do I care? A lot, apparently; if you didn't care you wouldn't say anything. How much do you say your time is worth, again? You must have donated hundreds of dollars worth of your caring to the mailing list over the past few weeks. Unfortunately, valuing the time of others in the same way, you've also cost the user community many thousands of dollars reading your strangely-embittered commentary. No one is forcing you to read anything. I never copy you on anythere, but here you are again Aside from the few weenies like yourself who think you know everything and would rather not hear the truth, I'm sure that there are many who are a bit more objective and value hearing another point of view. People want to know whats good and bad about using FreeBSD. Your Klan just paints a rosey picture about everything, so nothing you say can have any credibility. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Development Resources
In a message dated 10/16/04 5:27:37 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: Hello, After looking at the FreeBSD website and looking at docs all over the place, I havent found what I'm looking for, so I decided to mail this list. I am a software developer for Windows, and moving to FreeBSD has been very nice, especially since the tools to make software are completly free! My question is: Where can I find information on programming for FreeBSD? Things like how it differs from Windows, what it can and can't do, how to develop for X/KDE. I am good with C and C++, and know my way around gcc/make, but I don't know about system and 'net API calls that are specific to FreeBSD, and *NIX in general. If you can point me to a good website, that would help If you can point me to a (recent) book, that would be even better. Much thanks! Listen pal, there's are reason this stuff is free; figure it out for yourself! :D ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: dummynet
Why don't you guys stop torturing yourself and wasting $1000s worth of your time and get yourself some real bandwidth management software? Its cheaper in the long run. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Development Resources
In a message dated 10/28/04 2:49:23 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: Listen pal, there's are reason this stuff is free; figure it out for yourself! Dear troll, I notice that you conveniently omitted the smiley, you uptight loser. Relax a bit and get off the sauce. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: dummynet
In a message dated 10/28/04 12:52:14 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: Funny, I thought that's what Dummynet did. It seems that you wouldn't want to steer a user into a horribly overpriced closed-source rate-limiting solutuion when it's available for free in the OS. BTW: Nice email addr. ;) Ah, but its not really available for free, because the free ones don't work well, aren't supported and don't scale. Plus it seems that unless you value your time at $2./hr its already cost you more than the $800. to try to use the free stuff. Are you planning on completely rewriting it yourself using dummynet as the code base? What good is open source if the entire code base is nowhere near as good as what you can buy? You would really struggle with an inadequate open source solution rather than pay for something that works? And I wouldn't talk about email addresses, mr so liberal I can't function normally in society. AOL buffers the 99% of mails I have no interest in reading, I can just block the domains of lists I dont feel like dealing with at any given time without having to unsubscribe and subscribe, and it uses no disk space or bandwidth in the process. Its ideal (except for the darned reader). TM ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: GPL vs BSD Licence
In a message dated 10/28/04 4:49:28 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: I don't think that Allot modifies the Linux kernel. I wouldn't expect them to do so and I don't see an obvious reason why they should (*). Obviously some of their custom stuff needs to run inside kernel, but I rather think they enhance the kernel with some loadable modules or whatever (does Linux have KLDs?). Then you either know nothing about programming or nothing about their products. Do you think they do gigabit bandwidth management, with features not in the kernel, from user space? Plus, if they were using an unmodified kernel, why not provide the source? Put it on the machine. Whats the harm? A while back, I fast-read a post of Linus Torvalds to a mailing list saying why he thinks that binary-only enhancements to linux must be GPL licenced (and I believed the statemant was discussed on a FreeBSD-list also). His argument was that by using the kernel headers your work automatically becomes a derived work, thus it needs to be licensed under the GPL. I seem to recall the discussion was about nVidia's closed Modules use headers and are not GPLed, so clearly you're just plain wrong. Linus is just a big dope anyway, so who cares what he thinks? He's like Kerry. He thinks whatever is convenient for him to think at the time. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: dummynet
In a message dated 10/28/04 5:18:03 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED] I agree with some of that, but unless the person has the money to spend, then using dummnynet is acceptable. Not everyone can drop 10+ grand on a nokia firewall that has everything packaged into a nice gui. A commercial add-on for FreeBSD is $800. Half a weeks salary for a marginal programmer, and it actually works. Unless you live in Russia (or the Russian Federation or whatever the heck they call it now) and make $22/week I dont see the point of turturing yourself. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Troll (was: Development Resources)
The definition of a troll: A purposely stupid, inflammatory, or downright wrong article (closely related to flamebait). Its purpose is to get people mad and make them look stupid and gullible Definition of a Moron: Someone with no sense of humor. :) ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: dummynet
In a message dated 10/28/04 6:07:18 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: As far as being nowhere as good as you can buy, take a WatchGuard Firebox X1000 for example, they're pretty popular because they work. People that use them always tell me they prefer them to any *Nix based solution. By that statement, I know they've not really looked into that unit because the developers plainly state that it runs on a Linux hardened kernel. It terminates vpn connections, both ipsec and pptp, rate limits, nats and firewalls. All of the very same features you can do with Linux or FreeBSD using the appropriate packages. --- I never said anything about the O/S not being able to do it... works is a relative term. Most of the linux firewall/bwmgt boxes are just the same marginal stuff in the native O/S with a front end. Its better than nothing, but no better than dummynet, so no sense bringing them up. Allot's stuff runs on linux, etinc's stuff runs on both linux and freebsd. So it certainly can be done on un*x. The problem with dummynet is that once you do all the work and figure it all out, its still only marginally functional compared to something relatively inexpensive. So instead of buying the $3500 box that is everything you need, you've spend $800 on hardware, $2000 worth of time, and you still have something not nearly as good. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Troll (was: Development Resources)
In a message dated 10/28/04 6:42:28 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: Definition of a Moron: Someone with no sense of humor. :) No, it is more like someone who wastes everyone's time with useless junk just to irritate people. Try doing some real work. Coming from a guy who didn't know what sendmail was a few days ago, thats pretty darn scary. What real work are you engaged in Jerry? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: 4.10 - 5.2.1
In a message dated 10/27/04 6:49:14 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: Where can i find a step by step upgrade method (from FreeBSD 4.10 to 5.2.1)? Sorry if I seem lazy, but is somehow urgent (because a Plesk interface must be installed). Please post any link you think might be useful. Why would you want to do that? Its like going from a Mercedes to an Pinto. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: GPL vs BSD Licence
In a message dated 10/27/04 4:49:54 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: This is yet another example of the GPL license flaw. While any of the copyright holders of the Linux kernel could sue Allot, if they don't, it pretty much builds evidence that is going to help those that would argue that the GPL is uninforceable. There's been a couple of other GPL cases like this - of infringement that is being ignored. One of these days I'm going to have to gather up all these and write an article on it. ^- A problem with suing a company with a team of lawyers is that is will cost you a fortune, and what do you get? They have to agree not to sell it anymore? Its a pretty big project proving the validity of a license thats never really been seriously tested. You can't really claim damages if you have no intention of making money as a basic premise (and 3 times nothing is nothing). So they probably figure no one is going to do it. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: 4.10 - 5.2.1
In a message dated 10/27/04 4:55:08 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: Would you please use any form of quoting character to prefix the qouted lines of your reply? Any style would suffice. It's very hard to read your emails without. - Welcome to AOL my friend :) ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: GPL vs BSD Licence
In a message dated 10/27/04 12:59:24 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: If you buy a product what would you want ? A pretty box or pretty software ? Finishing the product is just marketing and trying to make a very pretty box to put the software in. When something is open source and you want to sell it you are forced to make it the best peace of code out there. Its what i call healthy competition. For me open source translates into If you think you can do better be my guest Finishing a product and making it closed source is just plain wrong. Its like stealing from the church basket. Every body shares something and you want to take it and keep it for your self. ^ I want : 1) a product thats finshed. Not with a long TODO list of basic features 2) a product that is bulletproof (or near so), that doesnt have only 25% of cases tested 3) a product that I don't have to spend 3 weeks of my time (@$300/hr) to get to the point that I can use it 4) a product where I have a contact that I can ask questions, and that I can expect to get obviously broken things fixed Any marginal programmer can write programs that do stuff. Getting people to be willing to pay for it is an entirely different level of talent and work. Its not just marketing. Marketing comes AFTER you have a product. Are people who have written custom GUI front ends for Linux stealing? They're not stealing, they are getting paid for the value that they've added. Are people that sell bottled water stealing? No one is forcing you to pay for water. But its been cleaned and nicely packaged and it fits in your cupholder, so you buy it. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Serious investigations into UNIX and Windows
In a message dated 10/26/04 12:24:06 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: If you think that administering a Windows server is so simple then answer the following test: How do you lock down an Exchange 5.5 server to prevent a spammer from using it as a relay. So who was the one who said either was easy? I said it takes a higher talent level to generally administer a un*x box than a windows box. I don't think that just because you can think of something thats not easy to do in windows makes any point at all. The fact that a un*x guy had to be called in to solve the problem says alot about the type of talent that is required to do most things that windows techs do. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: GPL vs BSD Licence
In a message dated 10/26/04 2:32:58 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: Actually a more interesting example is some of the Linksys routers do indeed use an embedded Linux along with Zebra as the routing engine. Ted Or Allot communications, who openly advertise the use of linux, but do not make source available to an obviously modified kernel.. I believe they claim that the GPL is optional. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Serious investigations into UNIX and Windows
In a message dated 10/26/04 10:07:10 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: [much snippage] Nonsense, if you ask me. For many reasons: a. Windows doesn't work nicely even for small networks most of the time. It's not the size of the network that matters. It's the nature of the network. Homogeneous, Windows-only networks will usually work somehow; not optimally, mind you, but they can be coerced into working. Heterogeneous networking environments, with many different types and versions of operating systems, are not so easy to use from Windows. the same can be said of Cisco based networks. Everything works better with products of the same make. Even NFS between different un*x boxes has issues. Integration is what separates the men from the boys, so don't complain. If it were easy most of us would be doing something else. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: GPL vs BSD Licence
In a message dated 10/26/04 2:26:45 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: Foundation, who is the copyright holder of the GPL license itself. In fact, the FSF advises authors to transfer copyright rights of their work to the FSF to avoid these problems. Ah, so your point is that people should transfer their copyrights to an organization dedicated to keeping the code free. Well, maybe they should,but that has nothing to do with which license is used. I think they both have it wrong. If you want to donate your code to the general community, make it available with no restrictions. The entire concept of here, use my crappy code but don't make any money off of it is totally lame. If someone takes it and doesn't give away the changes it doesn't diminish the original contribution. Its still there. Finishing a product is what has value. Anyone can write code that does this or that. Making it into something that someone is willing to pay for is what has value. And the more products that are available, the better off the community is. Even if they're not free. You still have the choice of paying for it or not. And you still have the original contribution to change as you please. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Serious investigations into UNIX and Windows
In a message dated 10/26/04 3:38:06 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: The fact that Cisco does something wrong doesn't somehow make it right for Windows. It's not a good excuse either. Its the way it is, and the way its always been. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: GPL vs BSD Licence
In a message dated 10/25/04 4:21:53 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: But equally important is the ability to join back forks, when/if some group finds the right solution to a problem. And that's where the GPL comes in: you can really think of the whole license as nothing more than a requirement to be able to re-join a forked project from either side. i don't really get what the gpl or bsd license has to do with rejoining forks. why shouldn't bsd licensed projects be able to refork in case... Because Juniper, for example, are perfectly free to decide against making their changes to the (in this case) FreeBSD code available anyone at all. For them, that may be a positive thing, because they don't have to open their work for the competition. But this is exactly what the GPL is aiming to avoid. --- The lack of foresight of the GPL is that, if Jupiter had no choice but to give away their work, then the work never would have been done, so even people willing to pay for it wouldn't have it. The GPL vs BSD issue is like liberal vs conservative. The liberal plaform sounds good and reasonable to those who don't understand the bigger picture. GPL is good if you're a programmer or hacker. But the companies that never put their corporate dollars into projects, because they can't make a profit from them, hurts the community in a different way. Products that would be available for sale aren't made available. People who can't spin their own don't get things that they need (which is why most companies use MS stuff). Most companies don't want source, they want stuff that works. FreeBSD is a perfect example of a thriving project with BSD licensing. Is FreeBSD a dead end? Is the community worse off because companies like Cayote Point and Emerging Technologies don't give the source to their products? No, because those products never would have been created if they were hindered by the GPL. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: ifconfig alias: File Exists
In a message dated 10/24/04 11:18:14 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: Is that new? You are right, that fixed it, but didn't think I had to do that before :( You get it because the guy who maintains ifconfig didn't have the foresight to realize the alias should imply a host mask, and also that the guy who coded the kernel code didn't think that assuming a host mask was reasonable. Welcome to open source. Love it and live with it. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED] To assume makes an ass out of u and me. Ok, that out of the way, the config you assume should be coded into ifconfig and kernel is not 100% going to be used all the time. In fact I have multiple nets and have multiple netmask assigned on the one machine. If you actually READ man ifconfig it states that this should be set to what you assume it should be. It helps when people don't attack things they don't fully understand cause for many it might be a person's first view at what you are bashing. Unfortunately also, many people aren't smart enough to get a second opinion or to try beyond there first try or someone person's like yourselfs comments. As for the assume thing, speak for yourself. Your implication that there should be no defaults is quite asinine. If it doesn't work with no netmask specified, then its broken. Its not unreasonable to assume that if no netmask is provided, then a host mask (for an alias) is intended. In the absence of a netmask, the only assumption thats reasonable is a host mask. There are lots of assumptions made by ifconfig. It assumes that you only want the interface to have one address (as if you submit an address to an interface that already has one it explicitly deletes the other). Its not unreasonable to assume that, nor would it be unreasonable to assume that the intention was to add an alias. It would certainly be safer. And I understand it a lot better than you do. In today's world, assuming the natural mask (which is what ifconfig has done since the beginning of time) is wrong most of the time. Just because someone back in the 1970s decided to do it that way doesn't make it correct. One of the basic properties of a default setting is that it should work. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Serious investigations into UNIX and Windows
In a message dated 10/24/04 5:54:56 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: I know more than a few people, small businessmen mostly, who have been completely screwed because their almost totally incompetent unix tech guy left the company. Ted wrote... For every small businessman screwed over this way there are ten times the number who have been screwed over by incompetent Windows tech guys. The point, Ted, is that you can easily find another incompetent windows tech, or even a good one to bail you out. With unix you're just screwed. Windows today is just as complex as any UNIX system. Sure, maybe a decade ago a peer-to-peer network of Windows systems your statement might have been true, but not today. You're also missing my point on this. You don't have to get into the guts of windows to make it work. You dont have to be a programmer to tweak all of the applications, in fact I know more than one windows tech who knows how to set things up but really has no idea what the settings mean. Yes you have to understand the applications to some degree. But to me, its a different level of skill to install and maintain applications in a unix-like environment. There's also less documentation, fewer resources, etc. So its more difficult to be proficient in unix than in windows. Ask a unix tech to install a windows application, or ask a windows tech to install a unix application. Which do you think has a better chance of success? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Serious investigations into UNIX and Windows
In a message dated 10/25/04 11:48:00 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: There's also less documentation, fewer resources, etc. So its more difficult to be proficient in unix than in windows. what are you talking about less documentation for Unix?!? What Unix are you referring to...Solaris...HPUX..AIX...BSD? I'm sorry to bite on this flame bait but i've been tracking your posts for some time now and I really don't understand where you get these ideas from? Let's see. The mailing list is freebsd-questions. Hmmm. I wonder. Perhaps the word logic in your domain name needs tweaking? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Serious investigations into UNIX and Windows
In a message dated 10/25/04 1:08:31 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: And this differs from your experience in the Windows world...how? :-) I'm not sure I understand your question. Rephrase or make it more specific, because answering to such a vague question is pointless. The inability of people to stay on point is as befuddling as it is entertaining. I think we all agree that you wouldn't let a windows tech touch your unix-like box, but you'd have no problem having a unix tech install a windows application. 'nuf said. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Serious investigations into UNIX and Windows
In a message dated 10/25/04 1:37:55 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: And this differs from your experience in the Windows world...how? :-) I'm not sure I understand your question. Rephrase or make it more specific, because answering to such a vague question is pointless. The inability of people to stay on point is as befuddling as it is entertaining. I think we all agree that you wouldn't let a windows tech touch your unix-like box, but you'd have no problem having a unix tech install a windows application. 'nuf said. Does this make you think at all? Does it worry you that the so-called technicians of the Windows world are often called to install, configure and run systems just because they know how to 'do' Windows? Is what you describe something that can be considered a disadvantage of the way UNIX works (making it pretty much obligatory to know what you are doing before doing it)? Or is it an advantage, after all? I haven't the foggiest idea what you're asking, but what I originally said what that, although unix may be better, there are reasons that people use Windows. Just because FreeBSD may be a better performer or perhaps more flexible, doesn't mean that its suitable for use any corporate environment. If you have a staff of people who know windows, you can't just move to FreeBSD and expect them to be able to administer it at the same level. The guy who originally posted was considering using Yahoo as evidence of FreeBSDs abilities. But the usefulness of an O/S is also a function of the talent that you have administering it. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: ifconfig alias: File Exists
In a message dated 10/19/04 3:51:33 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: # ifconfig fxp0 alias 200.46.204.9 ifconfig: ioctl (SIOCAIFADDR): File exists when I know for a fact that it hasn't been configured? you should use a netmask of 255.255.255.255 for ipv4 aliases. ifconfig fxp0 alias 200.46.204.9 netmask 255.255.255.255 Is that new? You are right, that fixed it, but didn't think I had to do that before :( You get it because the guy who maintains ifconfig didn't have the foresight to realize the alias should imply a host mask, and also that the guy who coded the kernel code didn't think that assuming a host mask was reasonable. Welcome to open source. Love it and live with it. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Serious investigations into UNIX and Windows
In a message dated 10/23/04 11:27:27 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: I have tried searching for this but i only get reports made by students and private programmers, i trust a programmer more than a large corporation any day but to show a person i know and convince him i need some serious investigations made by large corporations into comparing BSD and Windows systems in various areas. I would love to get some links that some of you have in your bookmarks on this. The person i'm trying to convince is a hardcore MS fan so i need real evidence of why BSD is better than MS products in server environments. Some friends of mine have told me that yahoo, msn and microsoft all use FreeBSD but until i can show him that and prove it to him that means nothing. Better for what? Every product is better at some things and weaker at others. You can argue that a BMW 325 is better than a 540 if you are concerned about gas prices, and its certainly more cost effective if your only use for the car is to go a short distance to work or the stores. Also, programmers have different criteria than non-programmers. Big companies are concerned with the ability to find people to administer their systems. There are more people around that can administer MS systems than unix, and it can be done with a lower level of talent. . A car enthusiast might prefer older, pre-computer cars because they're easier to tinker with. The same might be said for programmers. Programming types whine if they don't have source code, but source code is useless to people that don't know what to do with it (and its dangerous for those who only THINK they do). I think any high-level programmer who has used both unix and MS products is going to prefer unix for most things server-related, mainly because if it doesn't work just the way he wants he can likely fix it. On the other hand, there are more products available for MS, more vendors with supported products for certain, and if you're located in Moosebreath Montana and you need 40 guys to run an IT dept who know unix, good luck (unless you're willing to settle for a bunch of guys who know what YACC stands for and not much else). I know more than a few people, small businessmen mostly, who have been completely screwed because their almost totally incompetent unix tech guy left the company. FreeBSD is vastly better in a multitude of ways than an MS server on the same hardware, IF you have someone who knows what they're doing AND you can count on that guy hanging around. If not, you'll end up with a bunch of servers running poorly supported software that will run like the dickens until something happens, but that you won't be able to update, upgrade or repair. Of course there's no reason that you can't slap up a FreeBSD server until you're comfortable with it. I don't know of any law that says you have to decide between one or the other exclusively. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
7520 Chipset support in 4.x
Many of the new MBs from such tiny vendors as Dell and Supermicro are based on the 7520, and word is that FreeBSD 4.x doesn't support it. Is support forthcoming? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Freebsd and performance
In a message dated 10/20/04 9:19:30 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: If those things are taking a while to be there, does fbsd have any kernel patches like linux does to improve desktop performance? For example like: http://members.optusnet.com.au/ckolivas/kernel/ Lord, I hope not. One of the reasons I dislike Linux (and there are many) is how much independent, unofficial, hard-to-find, incompatible, and distribution specific development goes on. If you have patches that would benefit the project, why not submit them? Well, realize that linux patches are kludges that you don't want in the O/S proper. speeding desktop performance does things like bypass much of the stack, assuming that the box isn't going to route. Linux has performance features such as panicing if a single packet doesn't have headroom for a protocol header, because they don't want to do the check. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: The release of 5.3
I don't see how they can possible consider the Release of an O/S version when perhaps the most widely-available NIC (em) doesn't work. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Two Nics Two IP's same subnet what's going wrong ?
If both NICs are on the same network then you should set up bridging, not routing. Although some system may kludge it, the entire point of routing is that different network segments have different addresses. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: What version of FBSD does Yahoo run?
In a message dated 10/9/04 3:26:34 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of [EMAIL PROTECTED] Sent: Saturday, October 09, 2004 9:14 AM To: [EMAIL PROTECTED] Note that it also took quite a bit a beating to get them to admit that 1) They dont know the answer to the Subject or 2) Yahoo runs something quite different than what is available generally since they've substantially modified it. No, it is just I would be surprised if they didn't. Yahoo like any large company almost certainly has patentable ideas and a crew of lawyers reviewing everything. I would also expect they have a patent portfolio. Otherwise nothing would prevent some competitor ripping off their ideas and setting up a duplicate yahoo website. I would guess - since it is usual for this in most large companies - that some of these ideas are implemented in the FreeBSD they run. I don't work at Yahoo so I can freely speculate. And my speculations are founded on what is normal and usual for most larger companies. Nobody that works at Yahoo and actually knows the truth would be able to even speak hypothetically about what runs at Yahoo, as they would almost certaily be under an NDA. (something that is also normal and usual for most large companies) Ted - I don't see why Yahoo, or any other large company for that matter, would need or want to substantially modify the OS proper, as its a big win to *not* modify it so that you can run on whatever is the latest and greatest with minimal effort, you are certainly entitled to you opinion. Of course my point was that IF in fact you are right, and frankly I couldn't care less if you are or not, then the FreeBSD clan shouldn't be touting Yahoo as using freeBSD, any more than Ford can claim that some NASCAR driver drives a Mustang, if its been modified enough so that what they use is a completely different animal. If what Yahoo uses is based on FreeBSD, thats much different that using what everyone else does. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: FreeBSD Release Question
In a message dated 10/11/04 7:02:55 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: I hope you're not betting your business on these questions, because the reality is that 1) they're not very good questions and 2) the people who are answering them can't really know the answers. stable requires time, and since 5.2.1 and 5.3 are substantially different, I can't see how one can predict the level of stability a year from now. Following this logic any thing can be claimed to not be stable. -- uh, like yeah thats correct. Linux comes to mind ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: FreeBSD Release Question
In a message dated 10/9/04 6:25:49 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: 1. By Sep 2005, do you think 5.x performance will be optimized and be comparable to today's 4.x stable versions ? 5.3 is supposed to be stable, and it's expected to be on part with 4.x performance, and it's supposed to release before the end of the month. From what I've seen and heard, it looks like all that is going to happen. 2. By Sep 2005, do you think 5.x will be as stable as today's 4.x released versions ? Yes. I hope you're not betting your business on these questions, because the reality is that 1) they're not very good questions and 2) the people who are answering them can't really know the answers. stable requires time, and since 5.2.1 and 5.3 are substantially different, I can't see how one can predict the level of stability a year from now. You also didnt mention what your project is, so how can you expect anyone to comment on performance or stability? If you're developing a CD duplicator the answer is likely much different than if you are developing a networking product. If you can, do it on 4.x and move it to 5.x when you determine that it meets your needs. Don't bet the farm on the hopes and expectations of others. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: What version of FBSD does Yahoo run?
In a message dated 10/9/04 1:15:41 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: dang, how long is this thread gonna go on? Is it that important? I see a lot of good questions and equally good answers on this list, but I think this particular thread is starting to stoop beneath us all... Do you really read every thread? There are 100s of threads on here, many of them of no use, so why do you read them if you're not curious about it? It seems the most important question one could ask about FreeBSD is whether you should run 4.x or 5.x, and they always tell you to run 5.x because it suits the needs of Windbag River for guinea pigs. As long as you know you're a guinea pig, then you have your answer. I thought it was worth noting for the masses who unwittingly believe that a higher number release means better performance by default. Note that it also took quite a bit a beating to get them to admit that 1) They dont know the answer to the Subject or 2) Yahoo runs something quite different than what is available generally since they've substantially modified it. I yield the floor to the fat man in the toupee. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: What version of FBSD does Yahoo run?
In a message dated 10/9/04 12:56:40 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: I just hope that pounding packets through a socket and timing mySQL selects aren't the entirety of OUR team's arsenal. And it's not just this sentence although this is one of the most blatent. You are using verbage and terminology that clearly sets you in opposition to the rest of us, users and developers, of FreeBSD. If this isn't a challenge you don't know the meaning of the word. Maybe you think its a challenge because the words have teeth? Who is the rest of us in your estimation? Those under the thumb of wind river, or those of us trying to run small business who would prefer not to be bamboozled into using something new because you need free testers for your code? I monitored this list for months,and I never once heard any one of you tell anyone that 4.x was a better choice if running your business with the most efficient current solution was your goal. You don't care about the freeBSD community, you care about your own agenda, whoever you are. If you're not going to be honest with the community, then there's going to be a separation of you with the agenda and us with the need for honest answers to our questions so that we can run our businesses effectively. I love freeBSD. I have the skills to get my own answers as to the suitability of one OS or one version to another. Most people on this list don't. So don't steer them to 5.x when you know its not yet ready for prime time, because people rely on you to give good, honest answers in order to earn a living. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: What version of FBSD does Yahoo run?
In a message dated 10/8/04 2:42:11 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: Here is a thought. Why would they be running a pre-production release as a production server I have no idea what yahoo does, but I think it would be irrespondsible for them to attempt using 5.x on a production machine... I'm sure that Yahoo, like any large commercial enterprise, has a whole host of specific customizations that they have applied to FreeBSD, and their version of FreeBSD doesen't look like what we have, at least not where the good bits are. Why would they customize a beta, knowing that they'd just have to redo them when its released? I doubt they are that stupid. Also, If they've done substantial customization, then you really need to stop touting them as using FreeBSD, don't you, since they are not using whats available to everyone else. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: What version of FBSD does Yahoo run?
In a message dated 10/8/04 2:37:38 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: Kris and all, Sorry for the top post but would you quit feeding the trolls? Ted Mittelstaedt PS: TM, shut up and post some benchmarks proving your side of the argument. Not that we would believe them but you deserve to have to spend some time forging them up. Ah, so now anyone who questions your data is a Troll. Very convenient. The entire point of believability is the control, and the explanation of what the test actually tests. Thats the point of having a control, Ted. The test that was posted is not believable because it doesnt test anything that would actually happen in the real world. Do you buy a car because it hit 180 on the track? Is a car that can hit 190 but gets half the gas milage a better car? You guys are the ones making the claims that 5.3 is going to be so great. I just wonder how you come to that conclusion if you don't have any definitive tests. I dont have a release to test, so when its done I'll test it. - a relatively slow machine (a 1.7Ghz celeron with a 32-bit/33mhz fxp NIC running 4.9) pushes over 250Kpps, so why is your machine, with seemingly superior hardware, so slow? Because traffic is being generated from userland, not from within the kernel. - Actually my traffic generator is in userland too, of course. I guess I'm just a better coder than whoever wrote your little benchmark. Or maybe the benchmark is too busy calculating stats to do the work its supposed to be doing. Another variable in the test. For this workload, yes. It also seems that the gap has widened between UP and SMP performance in 5.x. Wasn't one of the goals of 5.x to substantially improve SMP performance? Yes, and it's ongoing. You don't see it on this workload, but there are other benchmarks (e.g. mysql select testing) that I don't have to hand at the moment, which show the smp benefits of 5.3 more clearly. This seems to show the opposite. No, it shows a small increase on SMP and a large increase on UP. Anyway, weren't you demanding an email ago that I produce benchmarks on UP systems, because no-one really uses SMP? You must be a democrat Kris, because you always spin what people say in a way such that is completely wrong when you say it. I said the 99% of us who don't use SMP, which is much different from no one uses SMP, isn't it? 1% of several million is not no-one, is it? Frankly, I didnt expect SMP performance to be so poor in 5.x since improving it is a stated goal. So I guess you recommend that anyone running a network server use a single processor? Are the gains in mySQL greater that the 40% loss in network performance? When mySQL is performaning so aptly, is the machine capable of handling a network load also? You (Kris) seem to think I'm asking you these questions, but I'm really not, but I guess I'm surprised you keep answering since you clearly don't have any of the answers. I'm just hoping someone does, somewhere. Because I don't see how you can develop an O/S without benchmarking your specific changes along the way. The folks at LINUX are guilty of building an O/S to suit their benchmarks. Its equally disturbing to implement theory without making sure that the theory works as expected. I just hope that pounding packets through a socket and timing mySQL selects aren't the entirety of your team's arsenal. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: What version of FBSD does Yahoo run?
In a message dated 10/8/04 2:25:52 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: You guys are the ones making the claims that 5.3 is going to be so great. I just wonder how you come to that conclusion if you don't have any definitive tests. I dont have a release to test, so when its done I'll test it. Why don't you download and install the latest 5.3 beta and test it? This is the only way to test 5.3, as there is no release yet. If you refuse to test it due to its beta status, then you should probably reconsider your challenges until 5.3 is stable and you can test it for yourself. If this is the case, then all you're doing now is making noise ... and, yes, trolling. - I haven't made any challanges. My point was that there are a lot of people making claims they have no ability to substantiate. And obviously I am correct. A guess a troll is anyone who questiong the powers that be. Must be a bunch of communists running the show here. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: What version of FBSD does Yahoo run?
In a message dated 10/6/04 6:47:04 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: On Wed, Oct 06, 2004 at 04:42:24PM -0400, Bigelow, Andrea L. wrote: Where's the documentation? I'd like to see this for myself. There is none, because Mr./Ms. TM4525 is making up his/her facts to suit their assertion. The last time this claim was made it was refuted and TM4525 promised to go away and check 5.3 performance. Kris -- Actually, Kris, it wasn't refuted, you said that the exceptionally poor performance was expected until 5.3 was released, and implied that anyone who expected good performance was making a fool of themselves. Search google groups for freebsd 5.2 performance woes and sort by date to see my test details and subsequent comments by Kris and the other FreeBSD Spin Doctors. My tests are very controlled, and my assertion is a result of exceptionally poor performance in the test. And no-one refuted my results. More like jockeying to save face. Nor did I promise to go away. I promised to test 5.3 and post the results. TM ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: What version of FBSD does Yahoo run?
In a message dated 10/7/04 10:17:47 AM Eastern Daylight Time, obscurity Kris writes: Well, it's vast :) Kris We're waiting..5.3 is in beta and ready for your tests. Other benchmarks show very good results compared to 4.x. Kris -- Quite a bunch of scientists on the FreeBSD team these days, eh? :) why don't you post some of these impressive benchmarks to substantiate your seemingly flimsy position? On a single processor system please, for the 99% of us who don't use SMP. Hopefully the only good reason to run 5.x won't be if you run 4 processor systems. TM ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: What version of FBSD does Yahoo run?
In a message dated 10/7/04 1:15:11 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: On Thu, Oct 07, 2004 at 12:41:28PM -0400, [EMAIL PROTECTED] wrote: In a message dated 10/7/04 10:17:47 AM Eastern Daylight Time, obscurity Kris writes: Well, it's vast :) Kris We're waiting..5.3 is in beta and ready for your tests. Other benchmarks show very good results compared to 4.x. Kris -- Quite a bunch of scientists on the FreeBSD team these days, eh? :) why don't you post some of these impressive benchmarks to substantiate your seemingly flimsy position? On a single processor system please, for the 99% of us who don't use SMP. Hopefully the only good reason to run 5.x won't be if you run 4 processor systems. Already done so. Kris Is it really too difficult for you to post a pointer or reference for those of us who don't have the time to spend our entire lives reading mailing lists archives? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: What version of FBSD does Yahoo run?
In a message dated 10/7/04 4:06:34 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: Here's one benchmark, showing UDP packet/second generation rate from userland on a dual xeon machine under various target loads: Desired Optimal 5.x-UP 5.x-SMP 4.x-UP 4.x-SMP 5 5 5 5 5 5 75000 75000 75001 75001 75001 75001 10 10 10 10 10 10 125000 125000 125000 125000 125000 125000 15 15 150015 150014 150015 150015 175000 175000 175008 175008 175008 169097 20 20 20 179621 181445 169451 225000 225000 225022 179729 181367 169831 25 25 242742 179979 181138 169212 275000 275000 242102 180171 181134 169283 30 30 242213 179157 181098 169355 That does show results for both single-processor (5.x-UP 4.x-UP) and multi- processor (5.x-SMP, 4.x-SMP) benchmarks. It may be that he ignored the table as soon as he read dual Xeon. I haven't seen this before. If I did, I would immediately ask: - What is the control here? What does your benchmark test? - Is this on a gigabit link? What are the packet sizes? Was network availability a factor in limiting the test results? - What does target load mean? Does it mean don't try to send more than that? If so, what does it show if you reach it? If you don't measure the utilization that it takes to saturate your target I don't see the point of having it. - It seems that the only thing you could learn from this test would be what is the maximum pps you could achieve unidirectionally out of a system. Why is that useful, since its hardly ever the requirement unless you're building a traffic generator? - a relatively slow machine (a 1.7Ghz celeron with a 32-bit/33mhz fxp NIC running 4.9) pushes over 250Kpps, so why is your machine, with seemingly superior hardware, so slow? - the test seems backwards. What you are doing in this test is not something that any device does. If you want to measure user-space performance, it has to include receive and transmit response, not just transmit. Perhaps it indirectly shows process-switching performance, but doesn't tell you very much about network performance, since transmit is much more trivial than receive in terms of processing requirements. When you transmit you know exactly what you have, when you receive you have to do a lot of checking and testing to see what needs to be done. When I test network performance, I want to isoloate kernel performance if possible. If you're evaluating the system for use as a network device (such as a router, a bridge, a firewall, etc), you have to eliminate userland from the formula. The interaction between user space and the kernel is a key factor in your benchmark that is absent in a pure network device, so its not useful in testing pure stack performance. Also, there is a significant problem with maximum packets/second tests. As you reach high levels of saturation, you often get abnormal processing requirements that skew the results. For example as you get higher and higher bus saturations the processing requirements change, as I/Os take longer waiting for access to the bus, transmit queues may fill, etc. Testing under such unusual conditions may inlcude abnormal recovery code to handle such saturations that would never occur with a machine under normal loads. A better way to test is measuring utilization under realistically normal conditions. Machines can get very inefficient if their recovery code is poor, but it may not matter since no-one realistically runs a machine at 98% utilization. Assuming that your benchmark does test something, Your results seem to show that a uniprocessor machine is substantially more efficient than an SMP box. It also seems that the gap has widened between UP and SMP performance in 5.x. Wasn't one of the goals of 5.x to substantially improve SMP performance? This seems to show the opposite. TM ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: What version of FBSD does Yahoo run?
Considering that its been well documented and admitted that 5.x is 1/3 the speed of 4.x at this point, do you really think they've migrated production boxes? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: When Unix Stops Being Fun - some advice
Some Advice, There are many things in life that seem like daunting tasks, some of them worthwhile, some not. But its the goal beyond the task that should be the deciding factor. Learning unix is not a reason. Its like saying you want to have children just for the sake of having them. Why do you want to learn unix? To enable yourself to start a business? To develop some great product idea? To enpower yourself to advance your career? Those are worthwhile reasons. There are lots of ways to occupy your mind. But its the ones with the really good reasons to learn it who are the best at it. Its also important to always remember (in life generally), that no matter how knowledgable you become, there will always be someone more knowledgeable, so don't be discouraged by others, or the fact that you are behind. Those others are the way you catch up, by listening to them, separating fact from bullshit, and advancing your own knowledge. The top of the bell curve is when you can spot the posers, the know-it-alls who really know nothing at all. Thats when you'll know you are on your way. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: When Unix Stops Being Fun
In a message dated 10/3/04 4:31:47 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: Excuse me while I shred it before the Secret Service comes knocking on my door... Is the secret service in charge of counterfiting now? (as you can see no formal education is required to be an SA) ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Freebsd 5.2.1 Performance Woes
In a message dated 9/30/04 12:03:05 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: Perhaps at this point you should go and do that, and avoid yourself any further embarrassment. I have read it, and I don't equate might be some regressions in performance to mean more than twice as slow. I also don't see any assurances that the performance of single processor systems is not being sacrificed in favor of improving multiprocessor performance. I'll be happy to test once its released. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Which FreeBSD For A Production System
In a message dated 9/30/04 2:04:35 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: Which release of FreeBSD is best for a production environment? I am aware of the different branches of development: CURRENT, STABLE, RELEASE and I *think* I understand the meaning of each from what I have read. Perhaps not since I am writing this question! But, what I would like to know is when I am setting up a production system, or desktop for that matter, which is considered *THE* most stable of the choices in versions. Is it in the 4.x branch, 5x etc... The most stable version of FreeBSD available today is 4.10. If your priority is to get something that will work, stay up, and not have to fiddled with, go with that. If you've got some time available, beta-testing 5.3 would be very helpful. In a matter of a few weeks, 5.3 is going to become -STABLE. It would be easier to stay up-to-date in the future if you go with 5.3. If its for a production system, then the only answer is 4.10. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: FreeBSD hardware specifications
In a message dated 9/29/04 8:44:14 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: On Wednesday 29 September 2004 04:12, annuar wrote: I'm interested on FreeBSD (download the 4.10) and would like to install it either on this machine or a new machine. If you are a new FreeBSD user, you might want to wait a week or two and download the 5.3 release. 5.x versions have been available for some time as new technology releases, but 5.3 will be a stable release. Lets be real. Anything with FreeBSD 5.x and stable in the same context is an oxymoron. It MAY be stable, IF it works on your motherboard, and IF you don't use a card that hasn't been tested, and IF there are no buglets in your bios and IF the stars line up in a pattern that looks like your grandmother. Try freebsd 4.10. Unless you're in some kind of hurry to go gray. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Freebsd 5.2.1 Performance Woes
While I was had a nice little test set up, I figured I'd test Freebsd 4.9 against 5.2.1 since I had fresh installs handy on separate drives. The simple test was as follows: Hardware: Celeron 2.4Ghz processor Dual onboard Intel (em) NICs, 32bit, 33Mhz bus Setup: Traffic Generator - FreeBSD System - Server I had it set up to route between the 2 interfaces (ie NOT bridged). I fired a unidirectional stream of UDP packets (so there was no return traffic from the server, which was discarding the packets) at a rate of ~100Kpps. The disturbing results: FreeBSD 4.9: 33% interrupt usage FreeBSD 5.2.1: 80% interrupt usage What has happened to the stack? ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Freebsd 5.2.1 Performance Woes
In a message dated 9/29/04 7:02:13 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: See the Early Adopter's Guide that was distributed with 5.2.1, or just don't worry about it and update to 5.3 which has vastly better network performance. Kris I'll post some numbers after trying it. But its pretty frightening to think that one release has such major changes. Sounds like (yet another) crapshoot. I'd have hoped that 5.whatever would be better than 4.anything simply as a matter of course. But 4.9 is slower than 4.7 and I can't help but worry that its just all downhill from here on. To think that its taken 18 months to get to be more than twice as slow as what was before is pretty discouraging. Nor can I have even the slightest bit of confidence that something so different than the previous release is going to be as stable as everyone seems to be claiming. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Device polling performance
In a message dated 9/25/04 4:24:07 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: The EVIDENCE is to the contrary, since it seems that a 2.4Ghz system will be saturated when bridging ~250Kpps with device-polling enabled, based on polling stats and userland benchmarking, even though the system claims to be 100% idle. Interestingly, its about the same with interrupt enabled. The POINT is that since there is no way to measure the performance, you've got a bunch of guys who think they've figured something out touting device-polling without having a clue what the performance advantages (or consequences) are, so it might as well be black magic, or snake oil, since you are as blind as a bat in your assessments. Hello, Please post your polling stats and userland benchmarking results. I would be very interested seeing them as I was thinking of moving to NICs that would benefit from polling. However, because you have EVIDENCE ... to the contrary, I may hold off. On the other hand, you do go on to say there is no way to measure the performance and you are as blind as a bat in your assessments, so also please post your test methodology. I need to make my decision on reliable, repeatable facts. Also, when you post, would you please wrap your lines to a shorter length? Not everyone on the list uses AOL Reader, like you. -- --- The evidence is a bit circumstantial in the absence of working tools, but here are some observations. There's also an assumption that the knobs associated with polling work as expected. Test machine is a 2.4Ghz celeron box with dual Intel NICs (em driver) on a 32bit, 33Mhz bus, running FreeBSD 4.9. Now I realize that a 133Mhz, 64bit bus is 8x faster and you certainly wouldnt use these NICs on a real network, but for the purpose of a control it doesnt matter, since both tests are on the same MB. Settings: HZ=1000 each_burst=512 max_burst=1000 user_frac=variable RXdescriptors (receive ring size) = 512 (Note that the burst never exceeded 100 at any time) I'm firing a controlled stream of 100K pps through the box (bridging). With only normal userland (idle) usage, the box happily goes about its way. Top shows 0-1.5% usage. I started a cpu intensive userland task (buildworld or something of the sort). The system started to lose packets with a user_frac setting of 78, which implies that the system requires about 22% of the cpu to successfully manage the task, assuming the knob works (it appears to). The same machine, with interrupts enabled, uses about 26%, according to top. HOWEVER, setting hz back to 100, with interrupts enabled the usage went down under 25%. Given that, it can be argued that there is less than a 5% bonus for polling, which makes a lot more sense than what some of the kooks have been saying. Of course the point here wasn't to prove the difference, which Im still not sure of, but the evidence certainly is that top doesn't properly account for CPU usage in device_polling mode.. I'd expect a small bonus, but nothing earth-shattering, as the machine still has to do the same amount of work. Its not like the machine is really servicing an interrupt for every packet, since controllers have hold offs so they don't generate interrupts on top of each other, and multiple events are regularly handled with a single interrupt. Polling gives the appearance of a machine happily going about its business no matter how much traffic you throw at it, but what happens is that you lose packets when it becomes overmatched, which never happens on a system with interrupts enabled before it goes into livelock. While livelock isn't a good thing, if it only happens occasionally, at least you aren't losing packets. Additionally, with a HZ setting of 1000 you're also introducing quite a bit of latency: additional_latency = up to 1ms in receive ring + transmit time for burst-1 frames. Increasing HZ further would reduce the latency, but adds more overhead, which slims the advantages and defeats the purpose of polling in the first place. The bottom line is that there isn't so much difference as to think that polling is going to save the day, but if you don't care about latency or losing packets it can be useful in allocating cpu cycles to user space, if thats your priority. TM ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Device polling performance
In a message dated 9/27/04 3:04:22 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: Mike at sentex.net previously wrote: Given a decent CPU, you wont see very much of a load average at all in the 200Kpps / 100Mb range. Note that load average and CPU usage are two intirely different things. You could have a huge amount of CPU usage with a load average hovering around zero and somewhat vice versa too - eg high load average without a great deal of CPU usage - though that would be less common. jerry Since device polling is entirely a kernel process (and userland load average has nothing to do with it), his statement would have been completely irrelevent if he were, in fact, talking about userland load average. Load average is virtually useless and shouldn't be part of any conversation originating after 1990. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Device polling performance
In a message dated 9/24/04 11:28:36 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: I thought I'd reword my question since no one seemed to understand the first time. Is there a way to measure CPU kernel/interrupt usage when device polling is enabled on 4.x systems? top and systat both show 100% idle all of the time. Hi, As long as all your interfaces support polling, you should see hardly see any interrupt usage at all, as that is the whole point of polling. You can allocate more or less CPU cycles to flinging packets around via various sysctl settings. See the polling man pages for more info ---Mike Thanks, but that doesn't answer the question. Since polling cycles don't seem to be shown under any usage category, how do you know what your system usage is when polling is enabled? It seems like a big negative to me. Tommy ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Device polling performance
In a message dated 9/25/04 10:17:27 AM Eastern Daylight Time, [EMAIL PROTECTED] writes: At 09:57 AM 25/09/2004, [EMAIL PROTECTED] wrote: Hi, As long as all your interfaces support polling, you should see hardly see any interrupt usage at all, as that is the whole point of polling. You can allocate more or less CPU cycles to flinging packets around via various sysctl settings. See the polling man pages for more info ---Mike Thanks, but that doesn't answer the question. Since polling cycles don't seem to be shown under any usage category, how do you know what your system usage is when polling is enabled? It seems like a big negative to me. Read the MAN page. There is a whole section there on a number of MIB variables that display various statistics around polling. 50% of the CPU cycles are allocated to the system by default. If that 50% is used up, it will show up in top under system processes in top. Given a decent CPU, you wont see very much of a load average at all in the 200Kpps / 100Mb range. ---Mike Ah, so the capacity of a FreeBSD router is 10 million packets per second, since 200K pps only uses .1 % of system resources. Kudos to the FreeBSD team for developing a stack that uses no resources. It seems beyond unreasonable that, with interrupts enabled, 55% of the system is used, and with polling, ~ zero. Since its clear you have no idea what you're talking about, perhaps if someone who actually does would like to pipe in it would be useful. It seems obvious that the system resource is not accurately monitored with polling enabled, which is what Im trying to get someone to admit, or to tell me when it was or will be repaired. TM PS: and please dont tell me to read the man page again. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Device polling performance
In a message dated 9/25/04 1:06:40 PM Eastern Daylight Time, [EMAIL PROTECTED] writes: It seems beyond unreasonable that, with interrupts enabled, 55% of the system is used, and with polling, ~ zero. Inconceivable! Erm...I do not think that word means what you think it means. It's probably easier to claim device polling works by black magic, then to explain just how much useless overhead is added when the CPU has to service tens of thousands of interrupts per second at very high packet rates. If you only need to bridge packets from one NIC to another, modern NICs will do almost all of the work (busmastering DMA, checksum offloading, etc) without needing any CPU time at all. If you've got good hardware, device polling means that the system is usually constrained by PCI bus throughput and the actual network transmission speed of the NICs themselves, not by CPU overhead. If you attempt to use crappy hardware (try digging up some 10Mbs ISA NICs), then the CPU will have to do a lot more work, and device polling will take up more than nearly zero CPU time. If you do routing, have firewall rules (particularly using dynamic stateful rules), or NAT (particularly running userland natd), you'll probably see significant CPU being used there. The EVIDENCE is to the contrary, since it seems that a 2.4Ghz system will be saturated when bridging ~250Kpps with device-polling enabled, based on polling stats and userland benchmarking, even though the system claims to be 100% idle. Interestingly, its about the same with interrupt enabled. The POINT is that since there is no way to measure the performance, you've got a bunch of guys who think they've figured something out touting device-polling without having a clue what the performance advantages (or consequences) are, so it might as well be black magic, or snake oil, since you are as blind as a bat in your assessments. cheers, TM ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Device polling performance
I thought I'd reword my question since no one seemed to understand the first time. Is there a way to measure CPU kernel/interrupt usage when device polling is enabled on 4.x systems? top and systat both show 100% idle all of the time. TM ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Device polling question - Freebsd 4.9
I'm a bit confused about device polling. I put a counter in em_poll(), and then enabled device polling for 10 seconds and read back the counter, and it was called 1.5million times. This is with virtually no network activity and HZ set to the default of 100. The docs seem to imply that the polling function gets called once per clock tick, but that doesn't seem to be the case at all. I'm trying to figure the reasoning behind mucking with the HZ value, which may affect some other things and also seems to defeat the purpose of trying to reduce overhead. With the HZ value at 100 the system easily passes 90K pps, but Im concerned about what variations occur when the system is under load, or if there are time-consuming activities (like extensive firewall rules). Also, when using device polling, even under heavy network load top shows the system to be almost completely idle. How can I gauge system usage when using polling instead of interrupts? Its difficult to compare the 2 methods when one of them can't be measured. TIA Tommy Mato ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
