A concentrated effort by enough ppl with fat pipes (DSL, T1/3, cable) would probably affect it quite a bit... repeated reloads and mailbombs (not that I would advocate such things - it would slow down my connection, too 8^D)
Might have some trouble bringing it all the way to it's knees... --
I started a new yahoo account to never use... 6 months later (never posting it anywhere or using it) I had ~10-20 spams/month (total over 100)... good stuff!
have to go with a decent SCSI solution on this one... You can't depend on ATA/66 or 100 for any sustained transfer at all. The protocol doesn't support disconnect, so one drive can hold up the channel while you wait for another. Physically, ATA/66 *cannot* sustain 60MB/s.
With 'older' UW SCSI hardware (2 1997/1998 9GB 7200RPM IBM drives) I can sustain ~12MB/s from each, and if I add in my 10krpm drive, I can sustain a total that essentially maxes out my 40MB/s UW SCSI link. If you *need* to keep near 60MB/s, U2W is really your only cost effective choice. Get 4 drives and a card... yeah, it'll run you a little $$$, but you actually will have the performace (striping the disk set, of course).
If you have a dedicated 100BaseT Ethernet link, you might be able to get 20MB/s but not 60... certainly not onto the same system as the drives (PCI 32b/33MHz is ~132MB/s max).
Right, there are boards from a number of manufacturers that have that - add another 32bits onto the transfer, and crank up the speed.
64/66 PCI is also helpful for this sort of thing, and would be a great increase. 64bit especially (included in Alphas, Netfinities, and a whole bunch of other things) makes it possible to mix older and newer cards on one bus. A 66MHz bus will get slowed down by a 33MHz card in it, since everybody has to play by the same rules. PCI-X is a better answer yet, but it tends to make devices a little more pricey.
ATA-100 is a waste of time. SCSI is a better answer.
Maybe I'll have to propose a Dual-PPC Netfinity, with high-speed internal I/O...
Or, you could just set up a caching RAID controller and that will help ease the disk I/O somewhat - how 'bout 8 10krpm U2W drives in a stripe set... that'd be nice. Not all you need is some higher bandwith through the chipset...
>If IBM took their S/390 technology and scaled it down, they would have a PC-killer device.
Get a small AS/400 - lots of I/O cabability, smaller footprint, still scalable. It's not exactly what you are looking for in a desktop - no AGP slot;-)
Yup, all sizes, from the new 250, all the way to the new 840 (that's a big horkin' machine)...
All the ones around here don't have the nice black covers on them... but maybe that's because I happen to spend time in a lab where we develop these things;-)
Having several VMs can be rather useful. You 'dedicate' a section of your mainframe to a specific task, give it a set amount of resources, and off you go. Running 41,500 copies of linux is just one of those "because I can" kind of things...
Actaully, if you are talking about massive transaction processing, the mainframes and other large systems (like the AS/400) are the best at what they do, and are actually rather cost-efficient. PC-style hardware has fairly awful I/O performance, which is what really matters for the big iron. Processor power is needed more for application serving, etc.. You can do a lot with big pipes. Sun systems have had great I/O capabilities for a while now, even with slower processors than PC equivalents, and AS/400 and S/390 go far beyond that. Not to mention that the reliability numbers on a 'frame are astounding... but when you a have a bunch of processors that do nothing but handle error recovery, it's no wonder why you don't see problems.
Most people might run maybe one or two VMs with linux, since that could provide a simple way to interface with the rest of the machine for some apps, but obviously, the power of the mainframe doesn't lie in its ability to run linux... it just happens to be a fun little add-on.
My user page has said that for a few days... I haven't had points in about 2 weeks... a little strange, isn't it? Methinks Rob is playing with something.
I'll second that... there's nothing like a nice bike ride (10-20 miles or so) or a good weightlifting session to get your mind off of other things so you can be more focused when you get back to them later.
If I haven't lifted or biked in a few days, my attention/focus span heads to about 5 seconds...
[] IA-64 is still on the way... there's still a bandwagon, but the wheels are a little rusty. [] oops... the PPC is probably the best chip out there in a lot of respects... [] Sure does... when you find the 21364, and have some review samples, I'll take some... oh, wait... [] Looks better than Merced - "My half-functional really hot chip is way better than your half-functional really hot chip... and it doesn't need quite as smart a compiler" [] Muhahahahaha (oops)
>Look at the Athlon, its not even native x86 it just emulates it.
If you want to get real technical, the PPro core (P-II, III, etc) emulates most of the x86 instruction set, too... the AMD happens to do it a little bit smarter, since it was redesigned from scratch (after fab processes had come a long way since 1994). When you know you will have the area and ability to make something, your design can be a lot less constrained.
As for the power consumption, if you have any extra 700 MHz 21264s to get rid of because they consume too much power, just toss 'em my way;-)
My linux webserver/firewall/NAT router is on a k6-2 500... runs great, and plenty of horsepower to compile large projects, too (mozilla, GCC for x-compile).
My other two systems are 1.5 yr old C-300a @ 450s... they run like a dream and barely get warm.
I'm waiting for the dual-athlon/duron/sledgehammer boards before I go all out on the next system (TNT-1 on a Celery@450 will have to hold me over on my gaming fix for now).
Agilent got all the stuff HP was famous for in the first place - top notch test equipment. Now all HP has is a bad PC line, a dying server series, and a (for you older guys out there) hot CEO.
Slashdot Mirror
please note the error in my last sentence... gigabit ethernet can do 20MB/s, 100BaseT can do 20Mb/s.... darn caps (and stupid me for mixing them up).
--
lucky you :)
--
A concentrated effort by enough ppl with fat pipes (DSL, T1/3, cable) would probably affect it quite a bit... repeated reloads and mailbombs (not that I would advocate such things - it would slow down my connection, too 8^D)
Might have some trouble bringing it all the way to it's knees...
--
I started a new yahoo account to never use... 6 months later (never posting it anywhere or using it) I had ~10-20 spams/month (total over 100)... good stuff!
--
have to go with a decent SCSI solution on this one... You can't depend on ATA/66 or 100 for any sustained transfer at all. The protocol doesn't support disconnect, so one drive can hold up the channel while you wait for another. Physically, ATA/66 *cannot* sustain 60MB/s.
With 'older' UW SCSI hardware (2 1997/1998 9GB 7200RPM IBM drives) I can sustain ~12MB/s from each, and if I add in my 10krpm drive, I can sustain a total that essentially maxes out my 40MB/s UW SCSI link. If you *need* to keep near 60MB/s, U2W is really your only cost effective choice. Get 4 drives and a card... yeah, it'll run you a little $$$, but you actually will have the performace (striping the disk set, of course).
If you have a dedicated 100BaseT Ethernet link, you might be able to get 20MB/s but not 60... certainly not onto the same system as the drives (PCI 32b/33MHz is ~132MB/s max).
Best of luck.
--
Yes, as OJ proved, te criminal and civil courts are a lot different...
--
Right, there are boards from a number of manufacturers that have that - add another 32bits onto the transfer, and crank up the speed.
64/66 PCI is also helpful for this sort of thing, and would be a great increase. 64bit especially (included in Alphas, Netfinities, and a whole bunch of other things) makes it possible to mix older and newer cards on one bus. A 66MHz bus will get slowed down by a 33MHz card in it, since everybody has to play by the same rules. PCI-X is a better answer yet, but it tends to make devices a little more pricey.
ATA-100 is a waste of time. SCSI is a better answer.
--
I'll have to check that out - thanks.
--
Hmmm... don't have that one yet 8^)
Maybe I'll have to propose a Dual-PPC Netfinity, with high-speed internal I/O...
Or, you could just set up a caching RAID controller and that will help ease the disk I/O somewhat - how 'bout 8 10krpm U2W drives in a stripe set... that'd be nice. Not all you need is some higher bandwith through the chipset...
--
>If IBM took their S/390 technology and scaled it down, they would have a PC-killer device.
;-)
Get a small AS/400 - lots of I/O cabability, smaller footprint, still scalable. It's not exactly what you are looking for in a desktop - no AGP slot
--
>I wonder if IBM needs beta testers (-: I'd re-wire my house if they sent me a demo unit.
I hope you have a nice big entry from the power company. The machine and AC might melt your puny 160A entry path 8^)
--
Right... or you could say that a cluster is just a lower-bandwith, less fault-tolerant mainframe (usually with more MIPS, though)...
--
>also, chicks dig it.
;-)
"Hey, baby.... check out my big iron."
--
Yup, all sizes, from the new 250, all the way to the new 840 (that's a big horkin' machine)...
;-)
All the ones around here don't have the nice black covers on them... but maybe that's because I happen to spend time in a lab where we develop these things
--
Having several VMs can be rather useful. You 'dedicate' a section of your mainframe to a specific task, give it a set amount of resources, and off you go. Running 41,500 copies of linux is just one of those "because I can" kind of things...
--
The S/390 provides the redundnacy... you can yank a few processors out of there, and you probably wouldn't even notice...
as for a extended power outage where your generator fails... well, that's another problem.
--
Actaully, if you are talking about massive transaction processing, the mainframes and other large systems (like the AS/400) are the best at what they do, and are actually rather cost-efficient. PC-style hardware has fairly awful I/O performance, which is what really matters for the big iron. Processor power is needed more for application serving, etc.. You can do a lot with big pipes. Sun systems have had great I/O capabilities for a while now, even with slower processors than PC equivalents, and AS/400 and S/390 go far beyond that. Not to mention that the reliability numbers on a 'frame are astounding... but when you a have a bunch of processors that do nothing but handle error recovery, it's no wonder why you don't see problems.
Most people might run maybe one or two VMs with linux, since that could provide a simple way to interface with the rest of the machine for some apps, but obviously, the power of the mainframe doesn't lie in its ability to run linux... it just happens to be a fun little add-on.
--
My user page has said that for a few days... I haven't had points in about 2 weeks... a little strange, isn't it? Methinks Rob is playing with something.
--
I'll second that... there's nothing like a nice bike ride (10-20 miles or so) or a good weightlifting session to get your mind off of other things so you can be more focused when you get back to them later.
If I haven't lifted or biked in a few days, my attention/focus span heads to about 5 seconds...
--
so...
[] IA-64 is still on the way... there's still a bandwagon, but the wheels are a little rusty.
[] oops... the PPC is probably the best chip out there in a lot of respects...
[] Sure does... when you find the 21364, and have some review samples, I'll take some... oh, wait...
[] Looks better than Merced - "My half-functional really hot chip is way better than your half-functional really hot chip... and it doesn't need quite as smart a compiler"
[] Muhahahahaha (oops)
--
>Look at the Athlon, its not even native x86 it just emulates it.
;-)
If you want to get real technical, the PPro core (P-II, III, etc) emulates most of the x86 instruction set, too... the AMD happens to do it a little bit smarter, since it was redesigned from scratch (after fab processes had come a long way since 1994). When you know you will have the area and ability to make something, your design can be a lot less constrained.
As for the power consumption, if you have any extra 700 MHz 21264s to get rid of because they consume too much power, just toss 'em my way
--
My linux webserver/firewall/NAT router is on a k6-2 500... runs great, and plenty of horsepower to compile large projects, too (mozilla, GCC for x-compile).
My other two systems are 1.5 yr old C-300a @ 450s... they run like a dream and barely get warm.
I'm waiting for the dual-athlon/duron/sledgehammer boards before I go all out on the next system (TNT-1 on a Celery@450 will have to hold me over on my gaming fix for now).
--
> all the hard-core techs will be naked by next year...
Thanks a lot... I just ate, doncha know!
--
>We should ask Jane!
>Jane? Oops, forgot to put in my earplug...
She doesn't want to talk to you now... she feels neglected...
--
Agilent got all the stuff HP was famous for in the first place - top notch test equipment. Now all HP has is a bad PC line, a dying server series, and a (for you older guys out there) hot CEO.
--