RAID0: yes, effectively. You can have multiple vdevs in the zpool. ZFS stripes data in the top-level vdevs, so there's your performance gain. RAID0+1: I don't believe so (you cannot nest vdevs) RAID1+0: yes. multiple mirror sets in the zpool, with striping among the top-level vdevs. RAID5: yes, effectively. Called raidz or raidz2 (for double parity). Unlike RAID5, the strip width is variable-sized, based on the size of the write and number of vdevs. Has the additional benefit of eliminating the RAID5 write-hole and eliminating readbacks for parity calculations. That alone is a big performance improvement in our environment.
By intentionally crashing IE when visiting a Google page?
By dropping packets containing 'oogle'?
By smearing Google's reputation with a astroturfing campaign?
By financing legal threats against them?
By buying legislation to outlaw 'renegade' technologies like PageRank, in favor of Microsoft's own intellectual-property-rights-respecting FairRank?
By throwing chairs at Google employees?
All of the above?
Because it damned sure isn't going to be by developing a new, innovative search and index technology and letting the market decide. The only market Microsoft respects is the one they own.
The filesystems going to be the hardest component of this. I know of no open-source fs that could handle this. I'm assuming this is all online storage, and there is no desire to nearline it to tape. Ideally, you'd want something that could contcatenate multiple LUNs (of RAIDed storage) without having to run through a volume manager. Nothing agaist volume managers, but it'd be another component to support. Looking at proprietary FSs, you've got CXFS from SGI, which could easily handle the PB requirement and plays nice on Linux. Sun's got QFS, which would max out at 1PB and could do the volume management bit easily. Linux support was a little flakey last time I used it, but it's a free download and evaluation, you could go get it right now.
IBM's SAN-FS would also meet the capacity needs and would have the advantage of providing nearline capability, if you're into that. Sun's SAM-FS is basically the QFS product with nearline-to-tape capability. Linux is only supported as a client OS there. Of course, if you buy the mantra that Solaris is 'open-source,' then that might not be an issue.
As for hardware with any of the above solutions, you're going to be looking at using multiple RAIDing disk enclosures of some kind. At a budget, probably SATA disks talking to the controller, and iSCSI to the host. FibreChannel to the host would be a little more costly, but might be worth it since iSCSI is just getting mature enough to be usable in production.
I've got a Steelcase Leap chair. First the office bought one for me, and then I would up buying one for home. Cost me $700, which may be more than you were looking to spend. Comes with a lifetime warranty on the mechanics (pneumatic cylinder, etc) and the cushions are user-replaceable. It is very adjustable...arm height, arm width, arm angle, chair height, seat depth, and seat edge. Plan on spending at least a week (took me 2) to get it set right for you. And then don't let other people muck with your settings. The best part is how the back of the seat is connected. Most chairs are kinda like a capital 'L' in profile, where the chair pivots on the lower left corner of the 'L.' On the Leap, the angle between the back and the seat increases as you recline. And the coolest part is that the seat - while remaining parallel to the floor - moves forward slightly while you recline. That lets me continue typing with an ergonomically correct arm position (and your elbows still on the armrests) as you change your degree of recline.
I don't have the reference pages right now, but....Most people's monitors are way too bright and have the contrast cranked up way too high. How do you know?
Contrast: Human eyes are better at seeing dark colored details on a light foreground, because of the way the eye adapts to the contrast range of a scene. Several optical illusions (Illusion 1) play on this fact. If you have eye fatigue problems using a white console w/ black text, you've got the contrast set too high. If green on black seems more comfortable than black on white, the contrast is too high
Brightness: Needs to be set according to the surrounding light level. The area behind and to the sides of the monitor should be about 75% as bright as the screen itself. I borrowed a light meter to check mine; place the sensor facing the wall, behind and to the side of your monitor to get a reading, then directly in front of your screen and adjust accordingly.
I've been having problems w/ eye fatige since the beginning of this year, and am getting to know my opthamologist fairly well. These are just notes I'm passing along from him as we try to get my workspace corrected.
This was about the most insightful post in this flame war. Let's remember three things
Most developers work on in-house projects.
For in-house applications, written for a single company and not to be distributed, the GPL'd QT seems perfectly acceptable. Since you don't intend to ever distribute the product, you are never hurt by the redistribution clause of the GPL.
Final decisions about important things are usually made my accountants, who measure everything in $.
So if the goal of UserLinux is to produce a Linux desktop suitable for businesses to build their critical software on, QT/KDE seems to be the better choice. The accountants will get nervous at the thought of spending a lot of man-hours on a project relying on GTK as a base, precisely because there is no single company (like TrollTech) they can write a big check to. Whether the company really needs the 'features' of the proprietary QT license, they will buy the license anyhow....it almost seems like an insurance policy to them.
Lest you think I'm being overly cynical: I've had a manager (an accountant) nearly cancel a nearly complete project, because it was built on PHP. She believed it should be developed in ColdFusion because that would cost money and hence be 'safer.' She had no clue about the underlying technology, but she had seen ColdFusion advertised in a magazine as the ideal platform for web development.
I will point out that the filesystems included in Novell's Netware product did include a deletion-recovery tool, accessible via the salvage command. My understanding was that Netware would not permantently delete a file until that disk space was needed for active data or until a timeout period expired.
Damned handy tool, too. We had IBM's TSM for our major backup operations, but for those "oops" moments, salvage was sure handy. I hope that the new Novell might consider implementing those features on existing linux filesystems, or at least contribute native linux implemenations of their filesystems.
I totally agree with this sentiment. Computers are a wonderful tool, especially for providing repetitive, adaptive practice so that students can improve their proficiency at a skill. But I think this money would be much better spent raising teacher salaries. Most teachers I know get paid so poorly that the really competent ones move on to better paying jobs.
"Also of note was the volume of OpenSource software in the box - OpenSSL/SSH, Apache, Samba, CUPS, Gimp-print, bash..."
If SCO is this dependent on OSS software, they are more vulnerable than I gave them credit for. A cohesive effort to remove support for Unixware might do them in. Sure...they have the source code and could re-add support, but it would be expensive for them, and they aren't going to be able to maintain that kind of payroll.
So how about it - how hard would it be to break support for SCO platforms? I mean, sure, I feel bad for existing Unixware users, but it would almost be doing them a favor to force them onto a modern OS
So either SCO was untruthful in its 10K filings, violating SEC regulations and its obligations to its stockholders and the financial community in large. Or....SCO is publically overstating its profitablity and assets (consequently boosting their stock price) while its executives dump all their holdings.
Either way, what they've done is illegal. The question for us - the open-source community - is finding someone with legal standing to make an issue out of it. Find two parties, one who would have been injured by the allegedly incorrect 10K filings, and other who would be injured by price pumping (recent buyers of SCOX). If each party sues for $XX million, someone would win and SCO should loose a big chunk of their relatively small cash assets, either by damage awards, SEC fines, or legal fees. Either way, it puts them out of our collective misery.
I concur with the poster above. Sounds like you want TSM for Space Management which supposedly works great in a multi-platform enviroment. Tivoli supports a number of NAS products, Win2k, and Linux. It needs a load of memory and CPU (IBM politely calls it a "CPU, memory, and I/O intensive application").
It is also a bitch to set up correctly, and will leave you crying in your beer for several weeks, unless you're a TSM wizard, in which case you are not getting paid enough. If you're working with a lot of critical data though, its the only way to fly.
Of course the government granted exclusive rights to Microsoft, in the form of copyright protection and software patents.
It has been suggested that when companies use such privileges illegally, they ought to lose them. It seems a logical enough proposition to me....
Of course, my lawyer friends think I'm nuts for believing this.
Slashdot Mirror
RAID0: yes, effectively. You can have multiple vdevs in the zpool. ZFS stripes data in the top-level vdevs, so there's your performance gain.
RAID0+1: I don't believe so (you cannot nest vdevs)
RAID1+0: yes. multiple mirror sets in the zpool, with striping among the top-level vdevs.
RAID5: yes, effectively. Called raidz or raidz2 (for double parity). Unlike RAID5, the strip width is variable-sized, based on the size of the write and number of vdevs. Has the additional benefit of eliminating the RAID5 write-hole and eliminating readbacks for parity calculations. That alone is a big performance improvement in our environment.
Because it damned sure isn't going to be by developing a new, innovative search and index technology and letting the market decide. The only market Microsoft respects is the one they own.
The filesystems going to be the hardest component of this. I know of no open-source fs that could handle this. I'm assuming this is all online storage, and there is no desire to nearline it to tape. Ideally, you'd want something that could contcatenate multiple LUNs (of RAIDed storage) without having to run through a volume manager. Nothing agaist volume managers, but it'd be another component to support. Looking at proprietary FSs, you've got CXFS from SGI, which could easily handle the PB requirement and plays nice on Linux. Sun's got QFS, which would max out at 1PB and could do the volume management bit easily. Linux support was a little flakey last time I used it, but it's a free download and evaluation, you could go get it right now.
IBM's SAN-FS would also meet the capacity needs and would have the advantage of providing nearline capability, if you're into that. Sun's SAM-FS is basically the QFS product with nearline-to-tape capability. Linux is only supported as a client OS there. Of course, if you buy the mantra that Solaris is 'open-source,' then that might not be an issue.
As for hardware with any of the above solutions, you're going to be looking at using multiple RAIDing disk enclosures of some kind. At a budget, probably SATA disks talking to the controller, and iSCSI to the host. FibreChannel to the host would be a little more costly, but might be worth it since iSCSI is just getting mature enough to be usable in production.
I've got my QuickCam Pro USB (dark focus ring) to work with the nw802 module and gnomemeeting. Color's a bit off (yellowed) though.
I've got a Steelcase Leap chair. First the office bought one for me, and then I would up buying one for home. Cost me $700, which may be more than you were looking to spend. Comes with a lifetime warranty on the mechanics (pneumatic cylinder, etc) and the cushions are user-replaceable.
It is very adjustable...arm height, arm width, arm angle, chair height, seat depth, and seat edge. Plan on spending at least a week (took me 2) to get it set right for you. And then don't let other people muck with your settings.
The best part is how the back of the seat is connected. Most chairs are kinda like a capital 'L' in profile, where the chair pivots on the lower left corner of the 'L.' On the Leap, the angle between the back and the seat increases as you recline. And the coolest part is that the seat - while remaining parallel to the floor - moves forward slightly while you recline. That lets me continue typing with an ergonomically correct arm position (and your elbows still on the armrests) as you change your degree of recline.
I don't have the reference pages right now, but....Most people's monitors are way too bright and have the contrast cranked up way too high. How do you know?
I've been having problems w/ eye fatige since the beginning of this year, and am getting to know my opthamologist fairly well. These are just notes I'm passing along from him as we try to get my workspace corrected.
This was about the most insightful post in this flame war. Let's remember three things
So if the goal of UserLinux is to produce a Linux desktop suitable for businesses to build their critical software on, QT/KDE seems to be the better choice. The accountants will get nervous at the thought of spending a lot of man-hours on a project relying on GTK as a base, precisely because there is no single company (like TrollTech) they can write a big check to. Whether the company really needs the 'features' of the proprietary QT license, they will buy the license anyhow....it almost seems like an insurance policy to them.
Lest you think I'm being overly cynical: I've had a manager (an accountant) nearly cancel a nearly complete project, because it was built on PHP. She believed it should be developed in ColdFusion because that would cost money and hence be 'safer.' She had no clue about the underlying technology, but she had seen ColdFusion advertised in a magazine as the ideal platform for web development.
I will point out that the filesystems included in Novell's Netware product did include a deletion-recovery tool, accessible via the salvage command. My understanding was that Netware would not permantently delete a file until that disk space was needed for active data or until a timeout period expired.
Damned handy tool, too. We had IBM's TSM for our major backup operations, but for those "oops" moments, salvage was sure handy. I hope that the new Novell might consider implementing those features on existing linux filesystems, or at least contribute native linux implemenations of their filesystems.
I totally agree with this sentiment. Computers are a wonderful tool, especially for providing repetitive, adaptive practice so that students can improve their proficiency at a skill. But I think this money would be much better spent raising teacher salaries. Most teachers I know get paid so poorly that the really competent ones move on to better paying jobs.
"Also of note was the volume of OpenSource software in the box - OpenSSL/SSH, Apache, Samba, CUPS, Gimp-print, bash..."
If SCO is this dependent on OSS software, they are more vulnerable than I gave them credit for. A cohesive effort to remove support for Unixware might do them in. Sure...they have the source code and could re-add support, but it would be expensive for them, and they aren't going to be able to maintain that kind of payroll. So how about it - how hard would it be to break support for SCO platforms? I mean, sure, I feel bad for existing Unixware users, but it would almost be doing them a favor to force them onto a modern OS
So either SCO was untruthful in its 10K filings, violating SEC regulations and its obligations to its stockholders and the financial community in large. Or....SCO is publically overstating its profitablity and assets (consequently boosting their stock price) while its executives dump all their holdings.
Either way, what they've done is illegal. The question for us - the open-source community - is finding someone with legal standing to make an issue out of it. Find two parties, one who would have been injured by the allegedly incorrect 10K filings, and other who would be injured by price pumping (recent buyers of SCOX). If each party sues for $XX million, someone would win and SCO should loose a big chunk of their relatively small cash assets, either by damage awards, SEC fines, or legal fees. Either way, it puts them out of our collective misery.
Of course, IANAL. But it sounds good, doesn't it?
I concur with the poster above. Sounds like you want TSM for Space Management which supposedly works great in a multi-platform enviroment. Tivoli supports a number of NAS products, Win2k, and Linux. It needs a load of memory and CPU (IBM politely calls it a "CPU, memory, and I/O intensive application"). It is also a bitch to set up correctly, and will leave you crying in your beer for several weeks, unless you're a TSM wizard, in which case you are not getting paid enough. If you're working with a lot of critical data though, its the only way to fly.
Of course the government granted exclusive rights to Microsoft, in the form of copyright protection and software patents. It has been suggested that when companies use such privileges illegally, they ought to lose them. It seems a logical enough proposition to me.... Of course, my lawyer friends think I'm nuts for believing this.