RE: Recommendation for 1000BASE-SX card?
-Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Erik Trulsson Sent: Friday, June 02, 2006 12:40 AM To: Ted Mittelstaedt Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED] Subject: Re: Recommendation for 1000BASE-SX card? On Thu, Jun 01, 2006 at 05:01:08PM -0700, Ted Mittelstaedt wrote: -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Chuck Swiger Sent: Thursday, June 01, 2006 8:24 AM To: [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Subject: Re: Recommendation for 1000BASE-SX card? [...] but I'm generally of the opinion that FreeBSD works just fine, most of the time, on most hardware, without any specific tweaking or tuning to be entirely usable. It does not. In reality, current versions of FreeBSD work better on current versions of hardware. FreeBSD has a terrible history of breaking things that used to work on old hardware, then when someone complains that something is broken, the developers in effect tell them their old hardware is crappy junk and to buy new hardware. Try running FreeBSD 6.X on a 80486 or Pentium system. FreeBSD 6.x works just fine on a Pentium system, as long as you have enough memory. Most Pentium 60's and Pentium 133's shipped from the factory with no more than 32MB of ram. That's only enough to load FreeBSD itself, not any applications. I'm not talking your souped up Pentium 200 with 128MB of ram in it. But, even those will roll over and die if you try to bring up a desktop like gnome or KDE on them. Way way too slow. Ted ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
RE: Recommendation for 1000BASE-SX card?
-Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Danial Thom Sent: Friday, June 02, 2006 10:08 AM To: Ted Mittelstaedt; Chuck Swiger Cc: [EMAIL PROTECTED] Subject: RE: Recommendation for 1000BASE-SX card? --- Ted Mittelstaedt [EMAIL PROTECTED] wrote: -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Chuck Swiger Sent: Thursday, June 01, 2006 8:24 AM To: [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Subject: Re: Recommendation for 1000BASE-SX card? Very well, let me put it another way: if your opinions about what's wrong differ from most other people, you might do better to rely on a discussion involving facts rather than opinions. Or, it could simply be that he's not doing what most people are doing, so he is going to run into trouble that most people don't run into. I mention this because some people regard their own opinions so highly that they don't seem to be aware that other approaches exist and might even prove effective. Like you? Clearly there are drivers that are well supported and drivers that aren't. There are people out there trying to run their businesses and you seem to want to pretend that everything is just peachy and that everything can be tweaked and tuned a bit to be usable. I don't know about either the OP or your situation(s), Then, pray tell, don't comment. Instead thank your lucky stars that you have not had to deal with that kind of problem. but I'm generally of the opinion that FreeBSD works just fine, most of the time, on most hardware, without any specific tweaking or tuning to be entirely usable. It does not. In reality, current versions of FreeBSD work better on current versions of hardware. FreeBSD has a terrible history of breaking things that used to work on old hardware, then when someone complains that something is broken, the developers in effect tell them their old hardware is crappy junk and to buy new hardware. Try running FreeBSD 6.X on a 80486 or Pentium system. FreeBSD 4.11 runs just fine on that hardware, if a bit slowly. But, I don't need speed to control my garden sprinklers. Now, it is true that sometimes backwards compatibility can hurt you, it can cause you to maintain interfaces and structures that conflict with support of new hardware, it can sometimes put you into situations that cannot be automatically resolved, thus you have to create a knob for the user to twaddle one way or another, depending on what hardware they have or what they want to do. It can suck off developer time to maintain old junk that only a few people use, instead of putting in support for new crap that a lot of people use. So there is a balance beam of too much backwards compatability and not enough of it. Microsoft is most definitely way far on the side of bending over backwards to support everything, but most people don't realize that FreeBSD is way far on the other side of sacrificing hardware support at the drop of a hat when people lose interest in it. That's true of some other platforms, such as Apple hardware and MacOS X, or even Sun/SPARC boxes, as well. YMMV. Total apples and oranges comparison, not relevant to anything. If you have specific problems or a FreeBSD-driver to Windows-driver performance comparison, providing #'s and enough details to reproduce would be helpful. That has been done with the Broadcom driver exhaustively in the PR database, there's at least a dozen PRs on problems related to that chip. However it has not resulted in much code to fix the problem, or even interest among committers to apply the fixes that have been posted. So no, I don't think that doing that is helpful at all. In fact, I really think the PR system has gotten pretty much broken these days, there's too many bugs and not enough people working on them, and more coming in every day. What is needed is some developers putting some time into knocking down the bugs in the PR database, but instead we have the foundation dumping money into funding students on projects like The Summer of Code which basically ends up creating a lot of half-finished efforts that may or may not eventually get integrated into the operating system at some point down the road. Nobody wants to fix other people's bugs, that's boring stuff, that is the one area of Open Source where commercial software companies have a leg up over us. A commercial company can find some starving programmer and pay him, then put a manager over him to keep jerking the paycheck string to keep him on task to do the icky programming. Open Source has real difficulty with the concept that some things in it are broken, rather ickely broken, and totally un-fun to work on, and the only way your going to get them fixed is by whipping some
RE: Recommendation for 1000BASE-SX card?
Its easy enough for commercial companies to fix the bugs if they need to use the broadcom drivers. There's just little incentive to donate the code back with this bunch of rude, incompetent clowns that have become the FreeBSD micky mouse club. I don't think it's that being the problem. I think the problem is that the engineers at places like HP and ASUS and such, know perfectly well the Broadcom and the Realtek and the other cheapo-crappy ethernet chipsets are garbage. But, I think they figure that they are not going to throw expensive programming time on solving the problems of those chips in software. I think they spend the expensive programming time on their high-end gear, which has the Intel chipset and the other good stuff, high end parts in it. There was a time when name brand companies like Dell, HP Gateway, Micron, etc. etc. made 2 lines of computers. Cheapo crappy desktop gear, and expensive high quality server gear. What I think ruined it is too many people pressing cheapo crappy desktop gear into use as servers, it was cutting into the high-end server market in a big way. So, the Dell's and the HP's of the world realized they needed to create server lines (and the motherboard manufacturers realized this too with motherboard lines) that were marketed as servers, but were a lot cheaper than their high end servers. This would allow them to package the exact same crappy desktop parts in a box marked as a server and costing twice as much, yet not as much as the really good quality server gear. And so that is what is going on these days. __ Ok, well we've blown the yahoo buffer so I have to crop. I'm not sure that its those corporate monsters making a conscious effort to rip people off. The market is uneducated. Managers at those companies don't know anything, and the engineers that design MBs are asian robots that just do schematics and make the chips work. People selecting products today are not engineers and have no idea now to test hardware; heck even Matt Dillon admits that he doesn't understand how the PCI bus works, and he's trying to design an operating system. Doesn't care either. Its all about the CPU. Which is silly, since putting a big, honking CPU on a box with a bad chipset or a cheap NIC devalues the CPU to the point that you might as well just get something cheap. Virtually no-one has any clue about the performance of their box. People are willing to spend any amount on their MB and CPU, and they they'll go out and buy a realtek ethernet card, or a 32-bit gig card to save a few $$$. Its mindless. Its so mindless I can't believe it. And even if you explain it to them, they still don't understand. Its like a bunch of women buying clothes. Costs more, must be better. Its just crazy. DT __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
RE: Recommendation for 1000BASE-SX card?
-Original Message- From: Danial Thom [mailto:[EMAIL PROTECTED] Sent: Friday, June 02, 2006 10:59 AM To: Ted Mittelstaedt; Heinrich Rebehn; [EMAIL PROTECTED] Subject: RE: Recommendation for 1000BASE-SX card? --- Ted Mittelstaedt [EMAIL PROTECTED] wrote: -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Danial Thom Sent: Thursday, June 01, 2006 6:38 AM To: Heinrich Rebehn; [EMAIL PROTECTED] Subject: Re: Recommendation for 1000BASE-SX card? --- Heinrich Rebehn [EMAIL PROTECTED] wrote: Danial Thom wrote: The intel cards that use the EM driver are the best performing cards in FreeBSD that we've tested. We've test cards made by the same company that use the broadcom controllers and the intel cards are substantially better (ie use less CPU passing the same amount of traffic). Be careful using on-board controllers. Usually vendors, for some reason, don't wire them to the pci-x bus. Most supermicro boards wire the em controllers to the 32bit/33mhz bus and the tyan and supermicro opteron boards we've tested wire the broadcoms to a shared 1x PCI-E, both of which will not only give you poor performance, but are not capable of running full gigabit rates. DT The Intel card would be an INTEL Pro1000MF, right? This would be quite expensive (~ EUR 430), but good performance and stability would warrant that. ATM, we are using the onboard controller (Broadcom BCM5704C wired to the pci-x bus). I did not have opportunity to do performance measurements, but we do have problems with our Linkpro 1000SX/1000TX converters, the 3rd of which has already died. That's why i want to give a PCI-X card with fiber interface a try. No, that would be the 1000MT, the MF is a fiber card I believe. They are about US$120. in the US. How do you know its wired to the PCI-X bus, since I don't believe that the controller has a way of reporting the way that the intel controller does? What MB do you have? Also keep in mind that the bge driver is a piece of crap; driver quality is a much more telling factor in these free OS's than the card in many cases. The EM and FXP are the only drivers worth anything (mainly because neither were written by mass-driver mill man Bill Paul). After having fixed bugs in the bge driver I must stress how wrong this statement is for the bge driver. Bill Paul may or may not have been associated with the bge driver, whether he was or not is immaterial since the bge driver is basically a port of the broadcom-supplied Linux driver, the code is Broadcoms mostly, with hunks of Broadcom code removed (like that dealing with the PHY's) when it was too difficult to port. (apparently) The quality of the Broadcom driver isn't Bill Paul's, it's Broadcoms. No, I can assure you that the reason the Broadcom chips work like crap under FreeBSD is not due to Bill Paul, it is because the Broadcom hardware iteself is pure, unadulterated, stinking, bull crap. It is crappy even under the supported operating systems like Windows, it's craptitude reaches new heights on the crap pile. Broadcom missed their calling as an ethernet chipset designer, they should have gone into making vacuum cleaners, as they would certainly be the suckiest ones in that business. Ted I'll disagree with you on the authoring issue (without commenting on the crappiness of the controller), because it is ultimately the responsibility of the programmer to work around the quirks and even the bugs in any given controller, and the simple fact is that BP does a half-assed job; certainly not the kind of job someone whose sole responsibility was to maintain a particular driver. All complex controllers are a b*tch to write drivers for, and the ability to seemlessly integrate working code into the OS to mask the quirks is what separates the men from the boys. Saying the driver stinks because the example code stinks is a cop-out. But I didn't say that. I said the driver stinks because the HARDWARE stinks. When I can take a Windows box with a Broadcom chip in it, that is exhibiting timeouts and slowness, unplug it from one brand of 10BaseT hub, and plug it into another brand of 10BaseT hub, and then plug my laptop into the first hub port that the Windows box was in, and have absolutely no problems, and have the Broadcom Windows box work perfectly in the second brand of hub, that is crappy hardware. It is not drivers, and no amount of twaddling with code in the driver will fix it. All sample code stinks. The sample code should be just that; an example of how to program the controller. Absolutely no, not at all. It is very easy to write a sample driver source that is full of unexplained magic numbers, in fact the Broadcom driver that I tweaked was broken precisely because one of the prior FreeBSD programmers who
RE: Recommendation for 1000BASE-SX card?
The starving programmer was an exaggeration used to illustrate a point, I was not seriously suggesting to go out and hire a bad programmer. But, when you buy cheap crappy hardware it is cheap because the manufacturer has hired less talented programmers among other things, and you can only expect something that works not that works well -- I think its often difficult to distinguish between what is crappy, because good code can make bad hardware look good and vice versa. All ethernet controllers were designed by idiots. My first success story (now I don't want to let on to who I really am so I'll be vague), was an ISA card by a major vendor that locked up regularly, and it had a hideous reputation as being a bad card. It was the only card of its kind, and I needed it badly. They gave me schematics and said that they had tried and tried but couldn't find anything wrong with the card. They had contracted out to some brainfarm to write a driver, and the thing was this beautiful self-contained scheduler (this is like MSDOS 3 mind you) with documented source, the whole deal. Well I tore it apart, simplified the code, got rid of all the soft interrupt passes and cleaned up all the memory management code. Now the card worked like a charm, didn't lock up, ran better than their spec and Mega-Billon$- company couldn't believe that some 23yo kid wrote a driver that a company they paid 100K to couldn't get to work. My point is that until someone writes a really good driver you never know if hardware is any good or not. Now some hardware is hopeless. I'm not sure that the broadcom controllers are that hopeless. But since the intel cards work well and are cheap, who's going to spend the time to pour over the broadcom driver and make it better? There's a ton of I/Os in there that can be streamlined. But who's gonna do it? Its sure not worth my time. DT __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
RE: Recommendation for 1000BASE-SX card?
-Original Message- From: Danial Thom [mailto:[EMAIL PROTECTED] Sent: Saturday, June 03, 2006 4:26 PM To: Ted Mittelstaedt; Heinrich Rebehn; [EMAIL PROTECTED] Subject: RE: Recommendation for 1000BASE-SX card? I think its often difficult to distinguish between what is crappy, because good code can make bad hardware look good and vice versa. All ethernet controllers were designed by idiots. My first success story (now I don't want to let on to who I really am so I'll be vague), was an ISA card by a major vendor that locked up regularly, and it had a hideous reputation as being a bad card. It was the only card of its kind, and I needed it badly. They gave me schematics and said that they had tried and tried but couldn't find anything wrong with the card. They had contracted out to some brainfarm to write a driver, and the thing was this beautiful self-contained scheduler (this is like MSDOS 3 mind you) with documented source, the whole deal. Well I tore it apart, simplified the code, got rid of all the soft interrupt passes and cleaned up all the memory management code. Now the card worked like a charm, didn't lock up, ran better than their spec and Mega-Billon$- company couldn't believe that some 23yo kid wrote a driver that a company they paid 100K to couldn't get to work. Musta been one of those Intel SatisFAXion cards. ;-) My point is that until someone writes a really good driver you never know if hardware is any good or not. Now some hardware is hopeless. I'm not sure that the broadcom controllers are that hopeless. But since the intel cards work well and are cheap, who's going to spend the time to pour over the broadcom driver and make it better? There's a ton of I/Os in there that can be streamlined. But who's gonna do it? Its sure not worth my time. Precisely!! ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Recommendation for 1000BASE-SX card?
On Thu, Jun 01, 2006 at 05:01:08PM -0700, Ted Mittelstaedt wrote: -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Chuck Swiger Sent: Thursday, June 01, 2006 8:24 AM To: [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Subject: Re: Recommendation for 1000BASE-SX card? [...] but I'm generally of the opinion that FreeBSD works just fine, most of the time, on most hardware, without any specific tweaking or tuning to be entirely usable. It does not. In reality, current versions of FreeBSD work better on current versions of hardware. FreeBSD has a terrible history of breaking things that used to work on old hardware, then when someone complains that something is broken, the developers in effect tell them their old hardware is crappy junk and to buy new hardware. Try running FreeBSD 6.X on a 80486 or Pentium system. FreeBSD 6.x works just fine on a Pentium system, as long as you have enough memory. FreeBSD 4.11 runs just fine on that hardware, if a bit slowly. But, I don't need speed to control my garden sprinklers. -- Insert your favourite quote here. Erik Trulsson [EMAIL PROTECTED] ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
RE: Recommendation for 1000BASE-SX card?
--- Ted Mittelstaedt [EMAIL PROTECTED] wrote: -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Chuck Swiger Sent: Thursday, June 01, 2006 8:24 AM To: [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Subject: Re: Recommendation for 1000BASE-SX card? Very well, let me put it another way: if your opinions about what's wrong differ from most other people, you might do better to rely on a discussion involving facts rather than opinions. Or, it could simply be that he's not doing what most people are doing, so he is going to run into trouble that most people don't run into. I mention this because some people regard their own opinions so highly that they don't seem to be aware that other approaches exist and might even prove effective. Like you? Clearly there are drivers that are well supported and drivers that aren't. There are people out there trying to run their businesses and you seem to want to pretend that everything is just peachy and that everything can be tweaked and tuned a bit to be usable. I don't know about either the OP or your situation(s), Then, pray tell, don't comment. Instead thank your lucky stars that you have not had to deal with that kind of problem. but I'm generally of the opinion that FreeBSD works just fine, most of the time, on most hardware, without any specific tweaking or tuning to be entirely usable. It does not. In reality, current versions of FreeBSD work better on current versions of hardware. FreeBSD has a terrible history of breaking things that used to work on old hardware, then when someone complains that something is broken, the developers in effect tell them their old hardware is crappy junk and to buy new hardware. Try running FreeBSD 6.X on a 80486 or Pentium system. FreeBSD 4.11 runs just fine on that hardware, if a bit slowly. But, I don't need speed to control my garden sprinklers. Now, it is true that sometimes backwards compatibility can hurt you, it can cause you to maintain interfaces and structures that conflict with support of new hardware, it can sometimes put you into situations that cannot be automatically resolved, thus you have to create a knob for the user to twaddle one way or another, depending on what hardware they have or what they want to do. It can suck off developer time to maintain old junk that only a few people use, instead of putting in support for new crap that a lot of people use. So there is a balance beam of too much backwards compatability and not enough of it. Microsoft is most definitely way far on the side of bending over backwards to support everything, but most people don't realize that FreeBSD is way far on the other side of sacrificing hardware support at the drop of a hat when people lose interest in it. That's true of some other platforms, such as Apple hardware and MacOS X, or even Sun/SPARC boxes, as well. YMMV. Total apples and oranges comparison, not relevant to anything. If you have specific problems or a FreeBSD-driver to Windows-driver performance comparison, providing #'s and enough details to reproduce would be helpful. That has been done with the Broadcom driver exhaustively in the PR database, there's at least a dozen PRs on problems related to that chip. However it has not resulted in much code to fix the problem, or even interest among committers to apply the fixes that have been posted. So no, I don't think that doing that is helpful at all. In fact, I really think the PR system has gotten pretty much broken these days, there's too many bugs and not enough people working on them, and more coming in every day. What is needed is some developers putting some time into knocking down the bugs in the PR database, but instead we have the foundation dumping money into funding students on projects like The Summer of Code which basically ends up creating a lot of half-finished efforts that may or may not eventually get integrated into the operating system at some point down the road. Nobody wants to fix other people's bugs, that's boring stuff, that is the one area of Open Source where commercial software companies have a leg up over us. A commercial company can find some starving programmer and pay him, then put a manager over him to keep jerking the paycheck string to keep him on task to do the icky programming. Open Source has real difficulty with the concept that some things in it are broken, rather ickely broken, and totally un-fun to work on, and the only way your going to get them fixed is by whipping some slave until they do the filthy task. People would rather spend the gold that they have on nice, pleasant projects that treat everyone nicely and look good on Resumes, and are not icky, nasty, uncomfortable things to do that make you late
RE: Recommendation for 1000BASE-SX card?
--- Ted Mittelstaedt [EMAIL PROTECTED] wrote: -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Danial Thom Sent: Thursday, June 01, 2006 6:38 AM To: Heinrich Rebehn; [EMAIL PROTECTED] Subject: Re: Recommendation for 1000BASE-SX card? --- Heinrich Rebehn [EMAIL PROTECTED] wrote: Danial Thom wrote: The intel cards that use the EM driver are the best performing cards in FreeBSD that we've tested. We've test cards made by the same company that use the broadcom controllers and the intel cards are substantially better (ie use less CPU passing the same amount of traffic). Be careful using on-board controllers. Usually vendors, for some reason, don't wire them to the pci-x bus. Most supermicro boards wire the em controllers to the 32bit/33mhz bus and the tyan and supermicro opteron boards we've tested wire the broadcoms to a shared 1x PCI-E, both of which will not only give you poor performance, but are not capable of running full gigabit rates. DT The Intel card would be an INTEL Pro1000MF, right? This would be quite expensive (~ EUR 430), but good performance and stability would warrant that. ATM, we are using the onboard controller (Broadcom BCM5704C wired to the pci-x bus). I did not have opportunity to do performance measurements, but we do have problems with our Linkpro 1000SX/1000TX converters, the 3rd of which has already died. That's why i want to give a PCI-X card with fiber interface a try. No, that would be the 1000MT, the MF is a fiber card I believe. They are about US$120. in the US. How do you know its wired to the PCI-X bus, since I don't believe that the controller has a way of reporting the way that the intel controller does? What MB do you have? Also keep in mind that the bge driver is a piece of crap; driver quality is a much more telling factor in these free OS's than the card in many cases. The EM and FXP are the only drivers worth anything (mainly because neither were written by mass-driver mill man Bill Paul). After having fixed bugs in the bge driver I must stress how wrong this statement is for the bge driver. Bill Paul may or may not have been associated with the bge driver, whether he was or not is immaterial since the bge driver is basically a port of the broadcom-supplied Linux driver, the code is Broadcoms mostly, with hunks of Broadcom code removed (like that dealing with the PHY's) when it was too difficult to port. (apparently) The quality of the Broadcom driver isn't Bill Paul's, it's Broadcoms. No, I can assure you that the reason the Broadcom chips work like crap under FreeBSD is not due to Bill Paul, it is because the Broadcom hardware iteself is pure, unadulterated, stinking, bull crap. It is crappy even under the supported operating systems like Windows, it's craptitude reaches new heights on the crap pile. Broadcom missed their calling as an ethernet chipset designer, they should have gone into making vacuum cleaners, as they would certainly be the suckiest ones in that business. Ted I'll disagree with you on the authoring issue (without commenting on the crappiness of the controller), because it is ultimately the responsibility of the programmer to work around the quirks and even the bugs in any given controller, and the simple fact is that BP does a half-assed job; certainly not the kind of job someone whose sole responsibility was to maintain a particular driver. All complex controllers are a b*tch to write drivers for, and the ability to seemlessly integrate working code into the OS to mask the quirks is what separates the men from the boys. Saying the driver stinks because the example code stinks is a cop-out. All sample code stinks. The sample code should be just that; an example of how to program the controller. Taking example code and forcing it into a driver results in a garbage driver. Which is also why the strategy of hiring some starving programmer is not a viable option,since the talent of the programmer is directly proportiate to the quality of the finished driver. The margin between marginal programmers (guys that can get something done that works) and a superior programmer is very large. We're not talking about welders here. There could be a 50% performance difference between 2 drivers for the same controller written by programmers with different talent levels. A couple of fewer or more efficient I/Os, a better way to allocate memory, how you flip the rings; all can make a tremendous difference, particularly when you get to the 100K+ iterations per second levels of traffic. DT __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ___ freebsd-questions
Re: Recommendation for 1000BASE-SX card?
Danial Thom wrote: The intel cards that use the EM driver are the best performing cards in FreeBSD that we've tested. We've test cards made by the same company that use the broadcom controllers and the intel cards are substantially better (ie use less CPU passing the same amount of traffic). Be careful using on-board controllers. Usually vendors, for some reason, don't wire them to the pci-x bus. Most supermicro boards wire the em controllers to the 32bit/33mhz bus and the tyan and supermicro opteron boards we've tested wire the broadcoms to a shared 1x PCI-E, both of which will not only give you poor performance, but are not capable of running full gigabit rates. DT The Intel card would be an INTEL Pro1000MF, right? This would be quite expensive (~ EUR 430), but good performance and stability would warrant that. ATM, we are using the onboard controller (Broadcom BCM5704C wired to the pci-x bus). I did not have opportunity to do performance measurements, but we do have problems with our Linkpro 1000SX/1000TX converters, the 3rd of which has already died. That's why i want to give a PCI-X card with fiber interface a try. The 3com 996-SX is somewhat cheaper, does anyone have experience with that one? Thanks for all your replies :-) Heinrich ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
RE: Recommendation for 1000BASE-SX card?
ATM, we are using the onboard controller (Broadcom BCM5704C wired to the just a hint: be really careful of what kind of broadcom-chip you'll get - some are could be not/bad supported by bge(4)/bce(4). (check the archives/PRs on that). ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Recommendation for 1000BASE-SX card?
--- Heinrich Rebehn [EMAIL PROTECTED] wrote: Danial Thom wrote: The intel cards that use the EM driver are the best performing cards in FreeBSD that we've tested. We've test cards made by the same company that use the broadcom controllers and the intel cards are substantially better (ie use less CPU passing the same amount of traffic). Be careful using on-board controllers. Usually vendors, for some reason, don't wire them to the pci-x bus. Most supermicro boards wire the em controllers to the 32bit/33mhz bus and the tyan and supermicro opteron boards we've tested wire the broadcoms to a shared 1x PCI-E, both of which will not only give you poor performance, but are not capable of running full gigabit rates. DT The Intel card would be an INTEL Pro1000MF, right? This would be quite expensive (~ EUR 430), but good performance and stability would warrant that. ATM, we are using the onboard controller (Broadcom BCM5704C wired to the pci-x bus). I did not have opportunity to do performance measurements, but we do have problems with our Linkpro 1000SX/1000TX converters, the 3rd of which has already died. That's why i want to give a PCI-X card with fiber interface a try. No, that would be the 1000MT, the MF is a fiber card I believe. They are about US$120. in the US. How do you know its wired to the PCI-X bus, since I don't believe that the controller has a way of reporting the way that the intel controller does? What MB do you have? Also keep in mind that the bge driver is a piece of crap; driver quality is a much more telling factor in these free OS's than the card in many cases. The EM and FXP are the only drivers worth anything (mainly because neither were written by mass-driver mill man Bill Paul). DT __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Recommendation for 1000BASE-SX card?
--- Heinrich Rebehn [EMAIL PROTECTED] wrote: Danial Thom wrote: The intel cards that use the EM driver are the best performing cards in FreeBSD that we've tested. We've test cards made by the same company that use the broadcom controllers and the intel cards are substantially better (ie use less CPU passing the same amount of traffic). Be careful using on-board controllers. Usually vendors, for some reason, don't wire them to the pci-x bus. Most supermicro boards wire the em controllers to the 32bit/33mhz bus and the tyan and supermicro opteron boards we've tested wire the broadcoms to a shared 1x PCI-E, both of which will not only give you poor performance, but are not capable of running full gigabit rates. DT To clarify, I'd recommend trying an intel PCI-X copper card with the convertor. The bge driver is garbage and is likely at least part of your problem with the convertor. DT __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Recommendation for 1000BASE-SX card?
Danial Thom wrote: --- Heinrich Rebehn [EMAIL PROTECTED] wrote: Danial Thom wrote: The intel cards that use the EM driver are the best performing cards in FreeBSD that we've tested. We've test cards made by the same company that use the broadcom controllers and the intel cards are substantially better (ie use less CPU passing the same amount of traffic). Be careful using on-board controllers. Usually vendors, for some reason, don't wire them to the pci-x bus. Most supermicro boards wire the em controllers to the 32bit/33mhz bus and the tyan and supermicro opteron boards we've tested wire the broadcoms to a shared 1x PCI-E, both of which will not only give you poor performance, but are not capable of running full gigabit rates. DT The Intel card would be an INTEL Pro1000MF, right? This would be quite expensive (~ EUR 430), but good performance and stability would warrant that. ATM, we are using the onboard controller (Broadcom BCM5704C wired to the pci-x bus). I did not have opportunity to do performance measurements, but we do have problems with our Linkpro 1000SX/1000TX converters, the 3rd of which has already died. That's why i want to give a PCI-X card with fiber interface a try. No, that would be the 1000MT, the MF is a fiber card I believe. They are about US$120. in the US. Our building has fiber cabling, that's why i am looking for a fiber card. The 1000SX/1000TX converters that we use are just to unreliable. How do you know its wired to the PCI-X bus, since I don't believe that the controller has a way of reporting the way that the intel controller does? What MB do you have? It is a Tyan Thunder K8SD Pro S2882-D http://www.tyan.com/products/html/thunderk8sdpro_spec.html The spec says that the BCM5704C is connected to PCI-X Bridge A (64Bit,100MHz). Also keep in mind that the bge driver is a piece of crap; driver quality is a much more telling factor in these free OS's than the card in many cases. The EM and FXP are the only drivers worth anything (mainly because neither were written by mass-driver mill man Bill Paul). That really sounds bad. I wonder if others can confirm that. --Heinrich ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Recommendation for 1000BASE-SX card?
Danial Thom wrote: Also keep in mind that the bge driver is a piece of crap; driver quality is a much more telling factor in these free OS's than the card in many cases. The EM and FXP are the only drivers worth anything (mainly because neither were written by mass-driver mill man Bill Paul). No doubt you've written a truly remarkable replacement driver which your email is too small to contain. Unlike Fermat, however, you've presumably hidden the proof under the bridge where only trolls may go. I'm tempted to set a Followup-to: header to -chat or maybe /dev/null... -- -Chuck ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Recommendation for 1000BASE-SX card?
--- Heinrich Rebehn [EMAIL PROTECTED] wrote: Danial Thom wrote: --- Heinrich Rebehn [EMAIL PROTECTED] wrote: Danial Thom wrote: The intel cards that use the EM driver are the best performing cards in FreeBSD that we've tested. We've test cards made by the same company that use the broadcom controllers and the intel cards are substantially better (ie use less CPU passing the same amount of traffic). Be careful using on-board controllers. Usually vendors, for some reason, don't wire them to the pci-x bus. Most supermicro boards wire the em controllers to the 32bit/33mhz bus and the tyan and supermicro opteron boards we've tested wire the broadcoms to a shared 1x PCI-E, both of which will not only give you poor performance, but are not capable of running full gigabit rates. DT The Intel card would be an INTEL Pro1000MF, right? This would be quite expensive (~ EUR 430), but good performance and stability would warrant that. ATM, we are using the onboard controller (Broadcom BCM5704C wired to the pci-x bus). I did not have opportunity to do performance measurements, but we do have problems with our Linkpro 1000SX/1000TX converters, the 3rd of which has already died. That's why i want to give a PCI-X card with fiber interface a try. No, that would be the 1000MT, the MF is a fiber card I believe. They are about US$120. in the US. Our building has fiber cabling, that's why i am looking for a fiber card. The 1000SX/1000TX converters that we use are just to unreliable. How do you know its wired to the PCI-X bus, since I don't believe that the controller has a way of reporting the way that the intel controller does? What MB do you have? It is a Tyan Thunder K8SD Pro S2882-D http://www.tyan.com/products/html/thunderk8sdpro_spec.html The spec says that the BCM5704C is connected to PCI-X Bridge A (64Bit,100MHz). Also keep in mind that the bge driver is a piece of crap; driver quality is a much more telling factor in these free OS's than the card in many cases. The EM and FXP are the only drivers worth anything (mainly because neither were written by mass-driver mill man Bill Paul). That really sounds bad. I wonder if others can confirm that. I clarified this in a second post, sorry. I'd recommend trying a copper card with your converter. I've tested that MB and I don't believe the controller is connected to a 64/133Mhz bus. Its less than have the speed (ie twice the load) as an EM card in the PCI-X slot. You can, of course, connect a part to a pci-x buss at 32bits and 33Mhz. Nevertheless, the bge driver with the mobo is suspect (we've had to hack it a bit to get it to work properly with bridging at all). It doesn't seem to want to come up at a gigabit unless you give it an address. The driver is really junk, IMO. DT __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Recommendation for 1000BASE-SX card?
No, I use drivers that are good. I don't feel the need to fix all that is broken in an OS; a good engineer finds what works and what doesn't and adjusts accordingly. Intel controllers are better than broadcom controllers anyways, so simply avoiding broadcom controllers is the strategy of choice. DT --- Chuck Swiger [EMAIL PROTECTED] wrote: Danial Thom wrote: Also keep in mind that the bge driver is a piece of crap; driver quality is a much more telling factor in these free OS's than the card in many cases. The EM and FXP are the only drivers worth anything (mainly because neither were written by mass-driver mill man Bill Paul). No doubt you've written a truly remarkable replacement driver which your email is too small to contain. Unlike Fermat, however, you've presumably hidden the proof under the bridge where only trolls may go. I'm tempted to set a Followup-to: header to -chat or maybe /dev/null... -- -Chuck ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED] __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Recommendation for 1000BASE-SX card?
--- Chuck Swiger [EMAIL PROTECTED] wrote: Danial Thom wrote: Also keep in mind that the bge driver is a piece of crap; driver quality is a much more telling factor in these free OS's than the card in many cases. The EM and FXP are the only drivers worth anything (mainly because neither were written by mass-driver mill man Bill Paul). No doubt you've written a truly remarkable replacement driver which your email is too small to contain. Unlike Fermat, however, you've presumably hidden the proof under the bridge where only trolls may go. I'm tempted to set a Followup-to: header to -chat or maybe /dev/null... What exactly is wrong with all of you people anyway? Clearly there are drivers that are well supported and drivers that aren't. There are people out there trying to run their businesses and you seem to want to pretend that everything is just peachy and that everything can be tweaked and tuned a bit to be usable. The poor guy goes out and buys a big honking machine with dual opterons (probably) and he's getting half of the performance out of the box because he's using an ethernet controller thats a piece of crap, or a driver thats a piece of crap, or maybe both. If you're going to call people who tell the truth about things trolls, so be it, but all you're doing is showing your own delusional view of the world, or your own stupidity as an engineer, if you just think that everything works so damn well. DT __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Recommendation for 1000BASE-SX card?
Danial Thom wrote: [ ... ] I'm tempted to set a Followup-to: header to -chat or maybe /dev/null... What exactly is wrong with all of you people anyway? Why do people ask rhetorical questions? We're not you, evidently. Invert the question, and I get an answer that makes sense. Invert the answer? You don't understand the point, perhaps? Very well, let me put it another way: if your opinions about what's wrong differ from most other people, you might do better to rely on a discussion involving facts rather than opinions. I mention this because some people regard their own opinions so highly that they don't seem to be aware that other approaches exist and might even prove effective. Clearly there are drivers that are well supported and drivers that aren't. There are people out there trying to run their businesses and you seem to want to pretend that everything is just peachy and that everything can be tweaked and tuned a bit to be usable. I don't know about either the OP or your situation(s), but I'm generally of the opinion that FreeBSD works just fine, most of the time, on most hardware, without any specific tweaking or tuning to be entirely usable. That's true of some other platforms, such as Apple hardware and MacOS X, or even Sun/SPARC boxes, as well. YMMV. If you have specific problems or a FreeBSD-driver to Windows-driver performance comparison, providing #'s and enough details to reproduce would be helpful. Writing random flames about specific people is not helpful. [ ...EOT, at least for me... ] -- -Chuck ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Recommendation for 1000BASE-SX card?
--- Chuck Swiger [EMAIL PROTECTED] wrote: Danial Thom wrote: [ ... ] I'm tempted to set a Followup-to: header to -chat or maybe /dev/null... What exactly is wrong with all of you people anyway? Why do people ask rhetorical questions? We're not you, evidently. Invert the question, and I get an answer that makes sense. Invert the answer? You don't understand the point, perhaps? Very well, let me put it another way: if your opinions about what's wrong differ from most other people, you might do better to rely on a discussion involving facts rather than opinions. I mention this because some people regard their own opinions so highly that they don't seem to be aware that other approaches exist and might even prove effective. Clearly there are drivers that are well supported and drivers that aren't. There are people out there trying to run their businesses and you seem to want to pretend that everything is just peachy and that everything can be tweaked and tuned a bit to be usable. I don't know about either the OP or your situation(s), but I'm generally of the opinion that FreeBSD works just fine, most of the time, on most hardware, without any specific tweaking or tuning to be entirely usable. That's true of some other platforms, such as Apple hardware and MacOS X, or even Sun/SPARC boxes, as well. YMMV. If you have specific problems or a FreeBSD-driver to Windows-driver performance comparison, providing #'s and enough details to reproduce would be helpful. Writing random flames about specific people is not helpful. [ ...EOT, at least for me... ] -- -Chuck thanks for clarifying that you're more of a clown than an engineer, chuck. I'm sure that info will come in handy for others when you state your opinions. Here's the deal. When passing a controlled stream of packets (say 10Kpps) through a 5704 controller on the tyan MB, the cpu load is twice what it is when passing the exact same load through an intel card in a pci-x slot. Maybe the intel card is superior, maybe the driver is superior, but the bottom line is that the broadcom sucks in comparision. I don't care if the card works well in windows or something else; its not practical to re-write the driver, and the intel cards are cheap, so why give a rat's butt why the results are what they are; I just use something else. Perhaps you take exception to my comment about the author of the driver, but the fact is that the guy wrote 50 drivers from a template, he doesn't support them, he didn't optimize them for performance, nor did he thoroughly test most of them (since most have stupid little quirks, which is what happens when you write a driver from a template). That spells trouble to me. The intel drivers are supported on an ongoing basis and they were written specifically for the controller, so they ought to work better, even if the hardware is identical performance-wise. Its not a flame to say that something works better than something else. If you think that all hardware and all drivers just either work or don't, then you're not only not an engineer, but a total fool. Its sharing my experience. If you care not to listen, fine, but you're making more noise calling me names for reiterating my experience, and at least I'm conveying information that might be useful, unlike yourself. You obviously have nothing to contribute, so I don't see why you feel the need to pipe in with an uneducated opinion. DT __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
RE: Recommendation for 1000BASE-SX card?
-Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Danial Thom Sent: Thursday, June 01, 2006 6:38 AM To: Heinrich Rebehn; [EMAIL PROTECTED] Subject: Re: Recommendation for 1000BASE-SX card? --- Heinrich Rebehn [EMAIL PROTECTED] wrote: Danial Thom wrote: The intel cards that use the EM driver are the best performing cards in FreeBSD that we've tested. We've test cards made by the same company that use the broadcom controllers and the intel cards are substantially better (ie use less CPU passing the same amount of traffic). Be careful using on-board controllers. Usually vendors, for some reason, don't wire them to the pci-x bus. Most supermicro boards wire the em controllers to the 32bit/33mhz bus and the tyan and supermicro opteron boards we've tested wire the broadcoms to a shared 1x PCI-E, both of which will not only give you poor performance, but are not capable of running full gigabit rates. DT The Intel card would be an INTEL Pro1000MF, right? This would be quite expensive (~ EUR 430), but good performance and stability would warrant that. ATM, we are using the onboard controller (Broadcom BCM5704C wired to the pci-x bus). I did not have opportunity to do performance measurements, but we do have problems with our Linkpro 1000SX/1000TX converters, the 3rd of which has already died. That's why i want to give a PCI-X card with fiber interface a try. No, that would be the 1000MT, the MF is a fiber card I believe. They are about US$120. in the US. How do you know its wired to the PCI-X bus, since I don't believe that the controller has a way of reporting the way that the intel controller does? What MB do you have? Also keep in mind that the bge driver is a piece of crap; driver quality is a much more telling factor in these free OS's than the card in many cases. The EM and FXP are the only drivers worth anything (mainly because neither were written by mass-driver mill man Bill Paul). After having fixed bugs in the bge driver I must stress how wrong this statement is for the bge driver. Bill Paul may or may not have been associated with the bge driver, whether he was or not is immaterial since the bge driver is basically a port of the broadcom-supplied Linux driver, the code is Broadcoms mostly, with hunks of Broadcom code removed (like that dealing with the PHY's) when it was too difficult to port. (apparently) The quality of the Broadcom driver isn't Bill Paul's, it's Broadcoms. No, I can assure you that the reason the Broadcom chips work like crap under FreeBSD is not due to Bill Paul, it is because the Broadcom hardware iteself is pure, unadulterated, stinking, bull crap. It is crappy even under the supported operating systems like Windows, it's craptitude reaches new heights on the crap pile. Broadcom missed their calling as an ethernet chipset designer, they should have gone into making vacuum cleaners, as they would certainly be the suckiest ones in that business. Ted ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
RE: Recommendation for 1000BASE-SX card?
-Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Chuck Swiger Sent: Thursday, June 01, 2006 8:24 AM To: [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Subject: Re: Recommendation for 1000BASE-SX card? Very well, let me put it another way: if your opinions about what's wrong differ from most other people, you might do better to rely on a discussion involving facts rather than opinions. Or, it could simply be that he's not doing what most people are doing, so he is going to run into trouble that most people don't run into. I mention this because some people regard their own opinions so highly that they don't seem to be aware that other approaches exist and might even prove effective. Like you? Clearly there are drivers that are well supported and drivers that aren't. There are people out there trying to run their businesses and you seem to want to pretend that everything is just peachy and that everything can be tweaked and tuned a bit to be usable. I don't know about either the OP or your situation(s), Then, pray tell, don't comment. Instead thank your lucky stars that you have not had to deal with that kind of problem. but I'm generally of the opinion that FreeBSD works just fine, most of the time, on most hardware, without any specific tweaking or tuning to be entirely usable. It does not. In reality, current versions of FreeBSD work better on current versions of hardware. FreeBSD has a terrible history of breaking things that used to work on old hardware, then when someone complains that something is broken, the developers in effect tell them their old hardware is crappy junk and to buy new hardware. Try running FreeBSD 6.X on a 80486 or Pentium system. FreeBSD 4.11 runs just fine on that hardware, if a bit slowly. But, I don't need speed to control my garden sprinklers. Now, it is true that sometimes backwards compatibility can hurt you, it can cause you to maintain interfaces and structures that conflict with support of new hardware, it can sometimes put you into situations that cannot be automatically resolved, thus you have to create a knob for the user to twaddle one way or another, depending on what hardware they have or what they want to do. It can suck off developer time to maintain old junk that only a few people use, instead of putting in support for new crap that a lot of people use. So there is a balance beam of too much backwards compatability and not enough of it. Microsoft is most definitely way far on the side of bending over backwards to support everything, but most people don't realize that FreeBSD is way far on the other side of sacrificing hardware support at the drop of a hat when people lose interest in it. That's true of some other platforms, such as Apple hardware and MacOS X, or even Sun/SPARC boxes, as well. YMMV. Total apples and oranges comparison, not relevant to anything. If you have specific problems or a FreeBSD-driver to Windows-driver performance comparison, providing #'s and enough details to reproduce would be helpful. That has been done with the Broadcom driver exhaustively in the PR database, there's at least a dozen PRs on problems related to that chip. However it has not resulted in much code to fix the problem, or even interest among committers to apply the fixes that have been posted. So no, I don't think that doing that is helpful at all. In fact, I really think the PR system has gotten pretty much broken these days, there's too many bugs and not enough people working on them, and more coming in every day. What is needed is some developers putting some time into knocking down the bugs in the PR database, but instead we have the foundation dumping money into funding students on projects like The Summer of Code which basically ends up creating a lot of half-finished efforts that may or may not eventually get integrated into the operating system at some point down the road. Nobody wants to fix other people's bugs, that's boring stuff, that is the one area of Open Source where commercial software companies have a leg up over us. A commercial company can find some starving programmer and pay him, then put a manager over him to keep jerking the paycheck string to keep him on task to do the icky programming. Open Source has real difficulty with the concept that some things in it are broken, rather ickely broken, and totally un-fun to work on, and the only way your going to get them fixed is by whipping some slave until they do the filthy task. People would rather spend the gold that they have on nice, pleasant projects that treat everyone nicely and look good on Resumes, and are not icky, nasty, uncomfortable things to do that make you late for dinner. Ted ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Recommendation for 1000BASE-SX card?
We are running the S4882-D and it has the Broadcom GB dual adapter built in (recognized as Broadcom BCM5704C Dual Gigabit Ethernet, ASIC rev. 0x2003). It seems to work without problems and we've run both to two different networks (right now running only one). We are connecting at 100baseT so I don't have experience running at GB speeds. I'm not detecting any network bottlenecks at all and I've been watching very carefully due to stability problems. We are running 6.0 RELEASE amd64 SMP 4/8 CPUs 8 GB. I don't know if the 2882 is identical in architecture but if so, strongly recommend NOT using 6.0 and trying 6.1. We are panicing daily and experience about 4 minutes a day downtime. Never have had FreeBSD be so unstable in 7 years. On May 31, 2006, at 2:22 AM, Heinrich Rebehn wrote: Hi list, can anyone recommend a 1000BASE-SX ethernet adapter for PCI-X slot, that is well supported by FreeBSD-amd64?. I want to use it in a TYAN Thunder K8SD Pro (S2882-D) board. TIA, Heinrich Rebehn University of Bremen Physics / Electrical and Electronics Engineering - Department of Telecommunications - Phone : +49/421/218-4664 Fax :-3341 ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-amd64 To unsubscribe, send any mail to freebsd-amd64- [EMAIL PROTECTED] ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions- [EMAIL PROTECTED] ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Recommendation for 1000BASE-SX card?
The intel cards that use the EM driver are the best performing cards in FreeBSD that we've tested. We've test cards made by the same company that use the broadcom controllers and the intel cards are substantially better (ie use less CPU passing the same amount of traffic). Be careful using on-board controllers. Usually vendors, for some reason, don't wire them to the pci-x bus. Most supermicro boards wire the em controllers to the 32bit/33mhz bus and the tyan and supermicro opteron boards we've tested wire the broadcoms to a shared 1x PCI-E, both of which will not only give you poor performance, but are not capable of running full gigabit rates. DT --- YTResearch [EMAIL PROTECTED] wrote: We are running the S4882-D and it has the Broadcom GB dual adapter built in (recognized as Broadcom BCM5704C Dual Gigabit Ethernet, ASIC rev. 0x2003). It seems to work without problems and we've run both to two different networks (right now running only one). We are connecting at 100baseT so I don't have experience running at GB speeds. I'm not detecting any network bottlenecks at all and I've been watching very carefully due to stability problems. We are running 6.0 RELEASE amd64 SMP 4/8 CPUs 8 GB. I don't know if the 2882 is identical in architecture but if so, strongly recommend NOT using 6.0 and trying 6.1. We are panicing daily and experience about 4 minutes a day downtime. Never have had FreeBSD be so unstable in 7 years. On May 31, 2006, at 2:22 AM, Heinrich Rebehn wrote: Hi list, can anyone recommend a 1000BASE-SX ethernet adapter for PCI-X slot, that is well supported by FreeBSD-amd64?. I want to use it in a TYAN Thunder K8SD Pro (S2882-D) board. TIA, Heinrich Rebehn University of Bremen Physics / Electrical and Electronics Engineering - Department of Telecommunications - Phone : +49/421/218-4664 Fax :-3341 ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-amd64 To unsubscribe, send any mail to freebsd-amd64- [EMAIL PROTECTED] ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions- [EMAIL PROTECTED] ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED] __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
