I don't think so. BSD is a mature, intelligently designed OS, one that attempts to do things the right way, not force you into the way someone in Seattle thinks you should.
The complaints that Linux users have about BSD and vice versa are nothing like the complaints that folks have about NT. We'd be having Linux/BSD vs. NT wars.
One thing that the article missed is the pride of hackerdom. Nobody anywhere gets paid for maintaining a FAQ, or a mailing list, yet people do it everyday. Folks moderate newsgroups, release code, anything, just to make a name for themselves in the community. A box of software can be bought, the respect of your peers is priceless. And, as this very website shows, hackerdom can be it's own reward.
I think that the BSD v. Linux thing will go on, but it will be based on the relative merits of the two, not on the licensing differences.
I work for a major ISP (check my e-mail address) in the Network Security group, and this here is one of our biggest cans of worms.
I am not sure whose side I am on on this one.... When we have attacks underway, and customers down, law enforcement is a good thing, and we want the cops to do everything possible.
However, some of the courts are "rubber stamp", and all it takes to undergo "due legal process" is asking permission.
The one thing that no one has mentioned yet is enforceability. At one point, AOL had >50Gb dedicated to tree.exe, some stupid xmas windows program, in their mail spool. Our backbone, up to OC-48 (~10 Gbit/sec) in some places, carries a TON of information. What can you use to sniff that? If you can sniff that, what do you do to log the packets that you sniff? And then what do you do to analyze them?
What scares me is the lack of understanding on the part of law enforcement. You don't tap the 'Net like you tap a phone line, and they just don't get that. "No, no, I don't want everything on the backbone, just the e-mail from this one user." Well, the backbone is fiber across the entire continent, and you want me to filter on layer 7 information?
What we have to watch out for is legislation like that which recently passed in Australia, forcing ISPs to comply with technologically impossible court orders.
We have a.5TB netapp here. It is used for logs, home dirs, databases,....
I'd bet that with the dedicated hardware, a netapp with a GigE interface would spool data to the net as fast or faster than a {Sun|NT|Linux} box would, regardless of caching. The dedicated hardware, and an OS that fits on a floppy would just plain do a better job.
However, you are still running RAID[45]. These have adequate read performance, but cannot exceed the speed of a single mirror. Since the server is to be primarily read dominated, RAID0+1 (striping and mirroring) may be a better choice. It still provides data integrity, and much better speed in this application. Although you need more disks to achieve a certain size volume, the controller hardware is cheaper, and generally more reliable.
With any decent stripe/mirror implementation, each mirror can satisfy read requests independently. Having 4 mirrors is like having 4 independent copies of the data. Need more bandwidth to the files, add another mirror....
Bottom line: If you want RAID4/5, and NFS or CIFS, the netapp is they way to go. It will be faster than any general purpose box with comparable equipment. If you want RAID0+1 or any other protocols, go somewhere else.
However, the original note specified GigE as the network connection with ~144Mbit active load. Just about anything can supply that load without a problem. Unless you are planning to up the network bandwidth to 500+mbit, this discussion is headed in the wrong direction.
It doesn't matter either way. It would if you were swapping off a CD, but...
Hard disks are CAV, Constant Angular Velocity. They spin at 7200, 3600, 10000, etc. RPM, regardless of the current head location. There are an identical number of sectors per cylinder, across the whole disk.
CDs, LaserDiscs (some), are CLV, Constant Linear Velocity. They spin faster at the center of the disk and there is a constant number of sectors per inch. That's why your HD is 7200RPM, and your CD is 32X max variable. Eric Brandwine
Sure, NT may win. But if it is going to, the playing field had better be level.
I started with Linux at 0.97pl3. I had to re-compile the kernel to change the IRQ for my bus mouse (37 mins on my 486). Things have gotten better since then. I remember when my roomate got the first Pentium in the dorm. PCI support for linux was miserable. If you FTPd to the machine, it would boot.
Linux lags. Without massive commercial support, new technology cannot have stable efficient linux support immediately. Granted.
But we now have 1 data point. With a machine above Linux's memory limit, with 4 CPUS (and linux SMP support lagging), and a poorly tuned web server, running a filesystem designed for another OS, NT wins.
Now lets see a few more points on the graph. Tune the web server. Do NFS testing. Try a machine that linux excels on. I would not be surprised to find out that NT _is_ faster on the big machine. But maybe nowhere else.
And then try that quad with gigs of RAM under Solaris. That'll give them an idea what Linux will be in the near future.
M$ has an advantage here that linux can never overcome. CIFS is anything but common. The "standard" is a Microsoft brainchild. Running Samba on linux is an unfair test. The semantics of SMB match those of a windows environment, and do not align at all with UNIX filesystems.
I would like to see NFS performance benchmarked. As much as can be said, NFS is the native network file system for Linux. I'd like to see how NT stacks up, regardless of hardware, when being pounded on by a network of Linux boxen.
I would like to see Linux based web clients. Maybe a million win95 machines can open a ton of connections to a web server. But what about a few clients with highly optimized stacks that can saturate an OC-12 or GigE on the server?
Maybe Linux does not play as well in the windows world as windows does. (I actually think it will win... How embarrasing, to be beaten at your own game!) But it is king in its own domain, where NT can barely operate.
I work for UUNET Infosec. There is another concern here. My Linux based web server will still be here tomorrow, with MY content on it. And the day after, regardless of who does what to it.
NT, IIS, and CIFS are a block of swiss cheese. There are so many holes, so many ways in.
Perhaps they should run a set of tests against a web server while it is under DoS (denial of service) attack, run some real world traffic across the link while the test is running. Benchmarks are pretty useless if the machine does not survive the test....
Sure, there are some/8s out there (e.g. MIT), but breaking them up is not a nice thing to do. Switching the backbone entirely to CIDR would cut our routng tables by a factor of 10. It would just require re-IPing everything that has an address. Everything. Not gonna happen this week.
And 2^32 is a high limit, sure, but not that high. However, exponential growth is a quick sucker. IPv6 provides enough addresses to give an IP to every grain of sand on every beach in the world. Even counting huge tracts of waste due to allocating contiguous blocks, we are not going to run out until we leave the planet. Heck, no one is at all scared of running out of ethernet MAC addresses....
Slashdot Mirror
I don't think so. BSD is a mature, intelligently designed OS, one that attempts to do things the right way, not force you into the way someone in Seattle thinks you should.
The complaints that Linux users have about BSD and vice versa are nothing like the complaints that folks have about NT. We'd be having Linux/BSD vs. NT wars.
One thing that the article missed is the pride of hackerdom. Nobody anywhere gets paid for maintaining a FAQ, or a mailing list, yet people do it everyday. Folks moderate newsgroups, release code, anything, just to make a name for themselves in the community. A box of software can be bought, the respect of your peers is priceless. And, as this very website shows, hackerdom can be it's own reward.
I think that the BSD v. Linux thing will go on, but it will be based on the relative merits of the two, not on the licensing differences.
Eric Brandwine
I work for a major ISP (check my e-mail address) in the Network Security group, and this here is one of our biggest cans of worms.
I am not sure whose side I am on on this one.... When we have attacks underway, and customers down, law enforcement is a good thing, and we want the cops to do everything possible.
However, some of the courts are "rubber stamp", and all it takes to undergo "due legal process" is asking permission.
The one thing that no one has mentioned yet is enforceability. At one point, AOL had >50Gb dedicated to tree.exe, some stupid xmas windows program, in their mail spool. Our backbone, up to OC-48 (~10 Gbit/sec) in some places, carries a TON of information. What can you use to sniff that? If you can sniff that, what do you do to log the packets that you sniff? And then what do you do to analyze them?
What scares me is the lack of understanding on the part of law enforcement. You don't tap the 'Net like you tap a phone line, and they just don't get that. "No, no, I don't want everything on the backbone, just the e-mail from this one user." Well, the backbone is fiber across the entire continent, and you want me to filter on layer 7 information?
What we have to watch out for is legislation like that which recently passed in Australia, forcing ISPs to comply with technologically impossible court orders.
Eric Brandwine
We have a .5TB netapp here. It is used for logs, home dirs, databases, ....
I'd bet that with the dedicated hardware, a netapp with a GigE interface would spool data to the net as fast or faster than a {Sun|NT|Linux} box would, regardless of caching. The dedicated hardware, and an OS that fits on a floppy would just plain do a better job.
However, you are still running RAID[45]. These have adequate read performance, but cannot exceed the speed of a single mirror. Since the server is to be primarily read dominated, RAID0+1 (striping and mirroring) may be a better choice. It still provides data integrity, and much better speed in this application. Although you need more disks to achieve a certain size volume, the controller hardware is cheaper, and generally more reliable.
With any decent stripe/mirror implementation, each mirror can satisfy read requests independently. Having 4 mirrors is like having 4 independent copies of the data. Need more bandwidth to the files, add another mirror....
Bottom line: If you want RAID4/5, and NFS or CIFS, the netapp is they way to go. It will be faster than any general purpose box with comparable equipment. If you want RAID0+1 or any other protocols, go somewhere else.
However, the original note specified GigE as the network connection with ~144Mbit active load. Just about anything can supply that load without a problem. Unless you are planning to up the network bandwidth to 500+mbit, this discussion is headed in the wrong direction.
Eric Brandwine
It doesn't matter either way. It would if you were swapping off a CD, but...
Hard disks are CAV, Constant Angular Velocity. They spin at 7200, 3600, 10000, etc. RPM, regardless of the current head location. There are an identical number of sectors per cylinder, across the whole disk.
CDs, LaserDiscs (some), are CLV, Constant Linear Velocity. They spin faster at the center of the disk and there is a constant number of sectors per inch. That's why your HD is 7200RPM, and your CD is 32X max variable.
Eric Brandwine
Sure, NT may win. But if it is going to, the playing field had better be level.
I started with Linux at 0.97pl3. I had to re-compile the kernel to change the IRQ for my bus mouse (37 mins on my 486). Things have gotten better since then. I remember when my roomate got the first Pentium in the dorm. PCI support for linux was miserable. If you FTPd to the machine, it would boot.
Linux lags. Without massive commercial support, new technology cannot have stable efficient linux support immediately. Granted.
But we now have 1 data point. With a machine above Linux's memory limit, with 4 CPUS (and linux SMP support lagging), and a poorly tuned web server, running a filesystem designed for another OS, NT wins.
Now lets see a few more points on the graph. Tune the web server. Do NFS testing. Try a machine that linux excels on. I would not be surprised to find out that NT _is_ faster on the big machine. But maybe nowhere else.
And then try that quad with gigs of RAM under Solaris. That'll give them an idea what Linux will be in the near future.
M$ has an advantage here that linux can never overcome. CIFS is anything but common. The "standard" is a Microsoft brainchild. Running Samba on linux is an unfair test. The semantics of SMB match those of a windows environment, and do not align at all with UNIX filesystems.
I would like to see NFS performance benchmarked. As much as can be said, NFS is the native network file system for Linux. I'd like to see how NT stacks up, regardless of hardware, when being pounded on by a network of Linux boxen.
I would like to see Linux based web clients. Maybe a million win95 machines can open a ton of connections to a web server. But what about a few clients with highly optimized stacks that can saturate an OC-12 or GigE on the server?
Maybe Linux does not play as well in the windows world as windows does. (I actually think it will win... How embarrasing, to be beaten at your own game!) But it is king in its own domain, where NT can barely operate.
I work for UUNET Infosec. There is another concern here. My Linux based web server will still be here tomorrow, with MY content on it. And the day after, regardless of who does what to it.
NT, IIS, and CIFS are a block of swiss cheese. There are so many holes, so many ways in.
Perhaps they should run a set of tests against a web server while it is under DoS (denial of service) attack, run some real world traffic across the link while the test is running. Benchmarks are pretty useless if the machine does not survive the test....
Sure, there are some /8s out there (e.g. MIT), but breaking them up is not a nice thing to do. Switching the backbone entirely to CIDR would cut our routng tables by a factor of 10. It would just require re-IPing everything that has an address. Everything. Not gonna happen this week.
And 2^32 is a high limit, sure, but not that high. However, exponential growth is a quick sucker. IPv6 provides enough addresses to give an IP to every grain of sand on every beach in the world. Even counting huge tracts of waste due to allocating contiguous blocks, we are not going to run out until we leave the planet. Heck, no one is at all scared of running out of ethernet MAC addresses....