You got the answer you were fishing for without buying the book. The top post should be modded -5 Lazy-SOB
I bought the book two hours ago, Einstein. And I'm betting that it doesn't have an answer to my original question; the book doesn't look large enough. So you're wrong. My post was not a blatant tech support request, no matter what you may believe. I think I explained myself well enough already, even though I didn't really have to.
I normally don't bother replying to ACs, but if you wanna slam me, at least have the balls to put your name behind your words.
Now, post your findings to mail list somewhere, so the next poor soul can benefit from your pain.
I will certainly do that. This problem drove me nuts. Everyone said it was something different. Not unusual for a thing which has lots potential points of failure.
I also posted an edited version of my original slashdot reply on my web site. Google should be by soon to pick it up. Every time I figure something weird like this out I put it up there. Get some decent referrers from google, too, so it's helping someone.
-B
Re:I'd buy the book if it could explain this...
on
Managing RAID on Linux
·
· Score: 2, Informative
What do you think it'll buy you, honestly?
Well, I had thought that my IDE controller was bad, the IDE drivers are wonky, the raid tools stuff was weird, whatever. I mean, I had two drives which both worked great when used by themsleves. I put them in a RAID pair, and I got errors. Turns out I had DMA disabled on one of them, but I was looking at Linux software RAID as the culprit. I thought buyiung dedicated hardware would isolate any problems. It was a last ditch, straw-grasping effort to tell the truth.
I'm actually a fan of Linux's software RAID1. No "special" drivers, I can use any kernel I want, easy to set up, minimal performance impact, and fairly transparent to use. Now that I know why I was getting errors, and that it wasn't anything to do with software RAID, I'm fine with it.
have you tried diddling with the dma/hdparm settings. I know some controllers have "issues" with dma
Turns out that the error isn't actually related to RAID at all. I mean, it is. The drives don't have errors when used outside a RAID setup. Put 'em back into a mirrored pair, and I start getting errors. But the problem is really caused by Linux's software RAID, per se.
You are exactly correct about the DMA stuff, though. Someone else suggested it, and I found out that it was in fact DMA. I had DMA enabled on one drive and not the other. Take the drives out of the RAID pair, and they don't individually show errors. Put them together, errors. That's why I thought RAID was the culprit (and why the book might help).
Thanks for the suggestions, BTW. Very much appreciated.
Nice one. You just managed to post a question on/. in the hope of eliciting technical help, qualifying it with an unconvincing statement that you might even buy the book.
Dig those knee-jerk accusations, man.
I've got a Linux RAID setup, it's been giving me errors for a while. I read the book review, and was wondering if maybe the book had any info about those errors, since no online source I could find did. After all, the problems are most definitely related to the drives being in a RAID pair because they don't have problems otherwise. So I composed a wool-gathering post about wondering how much detail could fit in 245-odd pages, and whether or not the book was worth it.
Then I read what I was about to post and judged it to be completly useless, uninformative, and uninteresting. So I added a question as to whether or not anyone had actually read the book, and could they tell me if it had info about the errors I was seeing. That was basically useless as well, so I pasted in an actual error (in order to be specific and get away from some lame "uh, I have RAID and it has errrors... can the book help?" question; it was also easier to copy-n-paste than explain what the error was), explained my situation, and said I'd buy the book if it could help me. Turns out it probably wouldn't be able to, which is exactly what I wanted to know.
Anyway, there was the rational behind my post. Anything else you'd like me to explain to you?
-B
Re:I'd buy the book if it could explain this...
on
Managing RAID on Linux
·
· Score: 2, Informative
Dude, that's hardware. Turn off the dma on your drives with hdparm.
You know what? The other drive in the RAID pair (/dev/hdd) had DMA off, while/dev/hdb had it turned on. I don't know why that was the case. Perhaps my late night fiddling resulting in some sort of fat fingering (wait... that sounded really bad). Anyway, I decided to do some tests by copying about 150MB of MP3s to my array while setting DMA to either on or off.
With DMA on/off (regardless of which drive has DMA on or off), I get the errors. With it set to off/off, I don't get errors, and the array is slower than a wounded prawn and a huge CPU hog (the copy takes around 50 seconds and the load avg hovers around 4.50). I don't care about slow since this is an NFS/Samba server and CAT5 is my bottleneck. The CPU load I do care about since the box does other things besides simply serve files. With DMA set to on for both drives, I also don't get the errors, which is very cool. The copy takes around 10 seconds and the load avg is about 0.70. All to be expected, since DMA gives quite a performance boost. But it's good to know I can turn it on.
Looks like my issue was with wacked DMA settings, and not the hardware going bad. So thanks for getting me to take another look! I probably ought to go buy the RAID book now...
When I've gotten that error it has meant that the drive itself is heading towards the great hardware graveyard in the sky. Since it's raid1 you should be able to simply put in a new/dev/hdb and all should be fine.
Well, data safety is the reason I have RAID. Trouble is, this is the replacement drive on/dev/hdb. The first had the same problem, which is why I was looking for controller/cable/driver issues.
I'm tempted to go buy a real RAID controller card and get away from software RAID. Problem is the Linux drivers are usually pretty strange. I like being able to upgrade my kernel, for instance.
-B
I'd buy the book if it could explain this...
on
Managing RAID on Linux
·
· Score: 3, Interesting
I've looked everywhere about why I keep getting these error messages on my Red Hat 7.3, 2.4.18-3 kernel RAID1 setup:
I've looked all over the place for the answer, google, mailing list archives, Usenet, local Linux friends, etc. and haven't been able to find a definitive answer. It's like nobody really knows what that error messages really means.
Newsgroups suggested bad cables, so I replaced those (twice, once with brand new cables bought specifically for the purpose). Some info suggested the drive or the drive's controller was failing, so I replaced it. Other info pointed to my IDE controller, so I installed a new one dedicated only to the RAID pair. I saw info that said the raid tools were to blame, and to see if the errors go away when the mirror is broken. No dice. Other info I found suggested that it was the IDE drivers in the kernel and that the messages were nothing to worry about unless I was seeing data corruption. I'm not seeing corruption so I'm left with this option.
If the book can shed some light on the error message voodoo one sees with Linux's IDE driver, then I'll buy it. I'd pay double what they're asking, even.
That depends on what you need the machines to do. When you factor in management costs for a cluster of full-fledged 1u PCs, blades are in no way too expensive.
colo space isnt that pricey nowadays, even downtown manhattan dcs are pretty cheap.
Again, not my experience. Most people I've seen have a rack or 12 and they don't want to buy more square footage. Some are in a unuversity setting where floorspace can be at a premium. Many realize that they can replace a half rack of 1Us with a half rack of 3 times the server power.
you would be hardpressed to financially justify forking out that kind of money instead of getting a 1u dual p3 tualatin with 2 gbs of ram
Not true. At the college where I work, we can't afford a lot of high-end 1U machines, in either replacement parts or management cost terms. With an RLX, we pay a lot up front for the chassis, but we can add blades very cheaply. And if we have RLX's management stuff can also accommodate requests like "I'd like 27 x86 servers for a class on distributed computing.. can you have it ready by next week?"
Those "little pc's" would be great for a bunch of dedicated servers in a compact space... I wonder if one could remove the CD drive and put a notebook HD in? That would be perfect
It's already been done, and done better than a stack of these little CD-sized guys. The RLX deals are pretty damn amazing. I've had occasion to see two different models in the past two years, and have been impressed each time. My favorite has to the be Transmeta-based blades, just because the consume like 9 watts when sitting idle. They're cool enough that you'd have a hard time telling they were powered on.
What makes something like an RLX chassis better than stacking in "little PCs" is that RLX has some very nice mgmt software that comes with the whole unit. Basically, you dedicate one blade to do mgmt stuff, and the rest (whether you have one chassis or ten) can all be managed by it. You can have all the blades sitting there blank, and remotely (and programmatically) boot up and then re-image any number of them with Windows or Linux, in any configuration you've set up. (The OS images are actually just tarballs of previously-installed operating systems you've set up and saved. So you can dedicate one blade to OS imaging duty, put Red Hat in whater config you want on it, upgrade the kernel or whatever and then push that tarball out to a "test blade" if you want to see how your apps runs.)
You also get more hardware with something like an RLX. The newer ones have dual fibre channel NICs, dual Gig Ethernet NICS, and a dedicated backplane network for "out of band" management, and an optional layer 2 switch for that chassis. That all means that you can make a cluster out of them really easily. And it means that you can do away with their hard drives, boot off the net and use network disk everywhere while still keeping them as "individual" servers. One more bonus: you don't have a cabling nightmare, and don't really need KVM for every server. They are also designed with heat output in mind. You can literally fill a 42U rack full of them (which is a total of like 330-something P3s) and still power it up. They're hot-swappable, too.
I don't work for RLX, I've just seen them up close a couple times (we're demoing one unit now, and will get another soon). If you are thinking of making a cheap cluster, or just want a lot of PCs in a little space withut a management headache, you might do well to look into RLX.
That's far too complicated for your average crackpot, don't you think? All I'm asking them to do it count and divide in order to find out of they are crackpots or not...:-)
And if Ford offered to replace the faulty seatbelt for free, would you still blame them and be all salty?
That depends on what damages I incurred, if any. If they offered to replace it before anything bad happened, then my estimation of Ford would go up quite a bit. But whatever my new love for Ford and their "doing the right thing" it still wouldn't change the fact that the original seatbelt was in some way sub-standard and needed to be replaced. Not a big deal, unless it was the latest in a long line of parts that needed replacing fr one reason or another.
Sheesh. I'm going to start Linux/Unix/FreeBSD/Sparc/OpenSourceSoftware bashing soon. Maybe that'll catch on.
Be my guest. You're allowed to hold whatever opinions you want. But you'd get more mileage out of constructive criticism rather than outright bashing, like I said eailer. You also wouldn't be the first to complain about the inadequacies of non-MS software. To paraphrase an old saying: All software sucks; some just sucks less.
Or maybe I'm just a dumbass like Boy Wonder down below claims. Ain't having choices grand?
The Unabomber's manifesto violates a very important law: The chances that a written work was authored by a crackpot increase with the percentage of completely capitalized words in the work.
I don't know if anyone else had come up with a similar law before I thought of it a number of years ago (thanks mostly to the brilliant work of none other than Ivan Stang), so I'll put a flag in it right now and call it Wee's Law of Tinfoil Hats.
I guess if I don't wear a seatbelt in may car and get injured in an accident...I should blame Ford?
You should if the seatbelt doesn't work. Hell, you might even be able to blame Ford is the seatbelts are too difficult to operate or hard to find. I once owned a car that wouldn't start unless the driver's seatbelt was latched. My newer cars don't have this "feature". Can I blame them for that?
Seriously, the MS bashing on this site is soooooooo lame.
Not really, no. It's incredibly easy, that I grant you. That makes for more bashing than necessary, much of it from an uninformed (or "partially formed") opinions. MS makes generally shoddy software, typically driven by a need for the cash generated by the upgrade mill rather than the needs of end users. Things like security are not a priority (or haven't been until recently) because of this, and they get a lot of flak for it. Rightly so, IMO. This makes it easy to come down hard on them.
Having said that, one should always try to use the right tool for the job. If that job is Grandpa's PC, then Windows is probably the right tool for it. If the job is my database server or my cluster, or even my mail server, Windows probably isn't. That's my opinion, and I'm completely allowed to call bullshit when I see people trying to use MS products for jobs for which they are ill-suited.
Put another way: using my personal experience/expertise to form an opinion (which may just come down harshly against MS) is never somthing I will think of as lame -- no matter what the venue. That line of reasoning leads one to suppose that MS is beyond the range of people's opinions, or somehow exempt from scrutiny. Such is not the case. (BTW, this is true of Linux, *BSD, Solaris, Oracle, et al. as well.)
If the opinions you see as bashing are obviously uninformed and useless, then call them as such, but don't label everything you don't particularly agree with as "lame MS bashing".
That doesn't match my experience - the Linux versions of the new patches were out within days of the Windows patches.
Thanks to a couple ex-Loki guys (Sam Latinga and Mike Philips, most notably) this is often the case. I'm sorry to question your experience, but things
weren't alwaysthis way. I recall waiting months before finally giving up on Linux T2 (I had purchased both Linux and Win32 copies).
Who's to say that the goodwill which keeps Linux patches coming will continue? If Sam goes on vacation for two weeks while a Win32 patch is released, what will Linux T2 players do? Play only on Linux servers until he gets back, that's what.
The reason Tribes2 isn't alive (or even worth a mention) is that Sierra killed (literally) Dynamix right after T2 was released. "Thanks for all the hard work and late nights... Now go home." Of course, Sierra still releases patches for Windows, but with Loki gone as well, the Linux patches lag months behind them (meaning Windows and Linux players can't play on the same servers). Consequently, T2 doesn't have the mod community that Tribes did and so can't depend on that to stay alive.
I was a huge Tribes fan, and waiting for a long time for Tribes2 to come out since it meant that I could finally dump Windows and use Linux for everything. After about a year I realized that wanting to use Linux for everyday gaming use was a pipe dream. I still play Wolfenstein, SimCity, Tribes2 and old ROMs and stuff, but to play PC games these days you either need WineX or a Win32 partition.
I was looking for ROM to old games (MOO, MOO2, Starflight) a while back, and recalled that I had a box of floppies that also had some old games on them. Hardly any of the 3 1/2 floppies were any good and I couldn't even read the 5 1/4 inch ones I found. That got me curious about what will happen to my meticulously ordered and cataloged CD-R/ROM collection.
While I was indulging my data storage daydreams, I came across a discussion board thread which talks about the various issues surround storing digital media (pictures, in this case). It was pretty intersting reading. I hadn't thought about gold-plated CDs before, and that sounds like a great idea as long as the hardware to read them exists for the duration of the media's shelf life. Even NASA has been having trouble in that area.
At first blush, I'd say the way to save all the images would be some sort of distrubuted filesystem, a la Freenet. Package an ASCII metafile with the ROMs file format info along with the actual image file and that should do it. Some sort of centralized system of making sure that at least N copies exist in "the wild" and the data could be reasonably safe. I'm oversimplifying, of course, but it occurred to me that data integrity and file formats might not be the only barriers to long-term data storage. Governments aren't especially data-friendly 100% of the time, either. If you really want to save data for all posterity, you have to protect it from yourself as well.
I can't see how it's even "quietly". Red Hat has had this info on their website for a while now (can't recall exactly when I saw it and submitted it to slashdot, but it was before last December). As for "microsoftish"? Well, you get what you pay for. Most of the apps I personally run are patchable from sources outside RH. At work, we tend to get patches from the "vendor": ssh from openssh.org, apache from httpd.apache.org, PHP from php.net, etc. in addition to ones from redhat.com/errata/. Likely, we'll shift that balance even more.
I guess Red Hat is being microsoftish by trying to make a profit (maybe someday),... or maybe it's the windowsupdate.com like ability to patch over the web.
They already do make a profit. Not a large one, but they are profitable. Is this move intended to generate more revenue? You bet. I can imagine them saying "Well, if you want/need guaranteed errata, then buy Advanced Server. If all you need is Linux, download an ISO and then after a year patch yourself for free." That's fine.
This also certainly ties into their Red Hat Update service, which already does network patching, same as Windows Update. (Forgive me if you were making that point; I missed the meaning of "from the web", and am construing "web" to mean "network".)
I think they're more Microsoftish than you may think, and I say 'right on!'.
As long as I have source, and they don't force me to deal with choices they've made, I'm with you. As soon as Red Hat starts trying to remove my liberties by trying to think for me, like MS has made a mint doing, I'm switching. But dump them for trying to make money selling services to those who need/want them? What could be wrong with that? I'd rather download free ISO from a company that only supports it for one year than have to pay up PBS style.
I personally thought the guy was full of it. But they were the ones that had to run around our new building balancing the AC system and adjusting ductwork. Perhaps that was an off-the-cuff remark, I don't know. I recall that I was going to look online to see if that was the case (because I could justify my monitors), but I never did.
In some cases, they have have even had to downgrade the airconditioning, because it was cooling the room too fast after the switch.
At a previous company, I had four machines in my office, three of which had 21" monitors. They made me get rid of a couple monitors because they couldn't balance the AC. One of the environmental servcies guys told me that a single 21" monitor is the equivalent of 6 people, as far as cooling the room is concerned. I was pretty astounded by that.
We had the same thoughts when reading that: He should most certainly either put up or shut up. I mean, he is, after all, getting what he's paying for. If he can't find a video player app that doesn't suck then he should write his own. Instead he whines about how bad every free app he's found is.
I admit that mplayer isn't the the most friendly install with the best UI, but if I cared enough to bitch about how it's laid out, I'd grab the source and effect some changes. It's easier to throw a fit than to fix things.
They are keeping some old CRT monitors around for just those kinds of reasons. There are also practical constraints (labs that do stuff with digital video, etc) where CRTs are always going to be installed.
I was mostly talking about the Dell workstations. They order 20 at a crack, all with flat panels, unless specifically ordered with a CRT. They also get them for server rooms. and whatnot.
Personally, I like my 21" Sony CRT at home much better than the 1800FP I'm using right now -- video, games, whatever, be damned. A CRT looks much better to me and is certainly easier on the eyes.
The only benefits are power use and desk space...two things that rank very low on the ladder of importance for me. I'm certain a LOT of other people feel the same way.
I can tell you that those two things (well, one does, anyway) rank pretty high on a large organization's list. For example, I can tell you that any new computers which come with monitors bought by UCSD's CS dept have to be ordered with LCD monitors now. The power savings are pretty big, even though it may take a while to phase in the new machines and their flatpanels. A couple friends in various other large companies have also seen this trend.
My guess is that Sony is merely catering to business needs and pressures and not thinking of home users as much.
Slashdot Mirror
I bought the book two hours ago, Einstein. And I'm betting that it doesn't have an answer to my original question; the book doesn't look large enough. So you're wrong. My post was not a blatant tech support request, no matter what you may believe. I think I explained myself well enough already, even though I didn't really have to.
I normally don't bother replying to ACs, but if you wanna slam me, at least have the balls to put your name behind your words.
-B
I will certainly do that. This problem drove me nuts. Everyone said it was something different. Not unusual for a thing which has lots potential points of failure.
I also posted an edited version of my original slashdot reply on my web site. Google should be by soon to pick it up. Every time I figure something weird like this out I put it up there. Get some decent referrers from google, too, so it's helping someone.
-B
Well, I had thought that my IDE controller was bad, the IDE drivers are wonky, the raid tools stuff was weird, whatever. I mean, I had two drives which both worked great when used by themsleves. I put them in a RAID pair, and I got errors. Turns out I had DMA disabled on one of them, but I was looking at Linux software RAID as the culprit. I thought buyiung dedicated hardware would isolate any problems. It was a last ditch, straw-grasping effort to tell the truth.
I'm actually a fan of Linux's software RAID1. No "special" drivers, I can use any kernel I want, easy to set up, minimal performance impact, and fairly transparent to use. Now that I know why I was getting errors, and that it wasn't anything to do with software RAID, I'm fine with it.
-B
-B
Turns out that the error isn't actually related to RAID at all. I mean, it is. The drives don't have errors when used outside a RAID setup. Put 'em back into a mirrored pair, and I start getting errors. But the problem is really caused by Linux's software RAID, per se.
You are exactly correct about the DMA stuff, though. Someone else suggested it, and I found out that it was in fact DMA. I had DMA enabled on one drive and not the other. Take the drives out of the RAID pair, and they don't individually show errors. Put them together, errors. That's why I thought RAID was the culprit (and why the book might help).
Thanks for the suggestions, BTW. Very much appreciated.
-B
Dig those knee-jerk accusations, man.
I've got a Linux RAID setup, it's been giving me errors for a while. I read the book review, and was wondering if maybe the book had any info about those errors, since no online source I could find did. After all, the problems are most definitely related to the drives being in a RAID pair because they don't have problems otherwise. So I composed a wool-gathering post about wondering how much detail could fit in 245-odd pages, and whether or not the book was worth it.
Then I read what I was about to post and judged it to be completly useless, uninformative, and uninteresting. So I added a question as to whether or not anyone had actually read the book, and could they tell me if it had info about the errors I was seeing. That was basically useless as well, so I pasted in an actual error (in order to be specific and get away from some lame "uh, I have RAID and it has errrors... can the book help?" question; it was also easier to copy-n-paste than explain what the error was), explained my situation, and said I'd buy the book if it could help me. Turns out it probably wouldn't be able to, which is exactly what I wanted to know.
Anyway, there was the rational behind my post. Anything else you'd like me to explain to you?
-B
You know what? The other drive in the RAID pair (/dev/hdd) had DMA off, while /dev/hdb had it turned on. I don't know why that was the case. Perhaps my late night fiddling resulting in some sort of fat fingering (wait... that sounded really bad). Anyway, I decided to do some tests by copying about 150MB of MP3s to my array while setting DMA to either on or off.
With DMA on/off (regardless of which drive has DMA on or off), I get the errors. With it set to off/off, I don't get errors, and the array is slower than a wounded prawn and a huge CPU hog (the copy takes around 50 seconds and the load avg hovers around 4.50). I don't care about slow since this is an NFS/Samba server and CAT5 is my bottleneck. The CPU load I do care about since the box does other things besides simply serve files. With DMA set to on for both drives, I also don't get the errors, which is very cool. The copy takes around 10 seconds and the load avg is about 0.70. All to be expected, since DMA gives quite a performance boost. But it's good to know I can turn it on.
Looks like my issue was with wacked DMA settings, and not the hardware going bad. So thanks for getting me to take another look! I probably ought to go buy the RAID book now...
-B
Well, data safety is the reason I have RAID. Trouble is, this is the replacement drive on /dev/hdb. The first had the same problem, which is why I was looking for controller/cable/driver issues.
I'm tempted to go buy a real RAID controller card and get away from software RAID. Problem is the Linux drivers are usually pretty strange. I like being able to upgrade my kernel, for instance.
-B
Jan 26 04:15:02 hostname kernel: hdb: dma_intr: status=0x51 { DriveReady SeekComplete Error }
Jan 26 04:15:02 hostname kernel: hdb: dma_intr: error=0x84 { DriveStatusError BadCRC }
I've looked all over the place for the answer, google, mailing list archives, Usenet, local Linux friends, etc. and haven't been able to find a definitive answer. It's like nobody really knows what that error messages really means.
Newsgroups suggested bad cables, so I replaced those (twice, once with brand new cables bought specifically for the purpose). Some info suggested the drive or the drive's controller was failing, so I replaced it. Other info pointed to my IDE controller, so I installed a new one dedicated only to the RAID pair. I saw info that said the raid tools were to blame, and to see if the errors go away when the mirror is broken. No dice. Other info I found suggested that it was the IDE drivers in the kernel and that the messages were nothing to worry about unless I was seeing data corruption. I'm not seeing corruption so I'm left with this option.
If the book can shed some light on the error message voodoo one sees with Linux's IDE driver, then I'll buy it. I'd pay double what they're asking, even.
-B
That depends on what you need the machines to do. When you factor in management costs for a cluster of full-fledged 1u PCs, blades are in no way too expensive.
colo space isnt that pricey nowadays, even downtown manhattan dcs are pretty cheap.
Again, not my experience. Most people I've seen have a rack or 12 and they don't want to buy more square footage. Some are in a unuversity setting where floorspace can be at a premium. Many realize that they can replace a half rack of 1Us with a half rack of 3 times the server power.
you would be hardpressed to financially justify forking out that kind of money instead of getting a 1u dual p3 tualatin with 2 gbs of ram
Not true. At the college where I work, we can't afford a lot of high-end 1U machines, in either replacement parts or management cost terms. With an RLX, we pay a lot up front for the chassis, but we can add blades very cheaply. And if we have RLX's management stuff can also accommodate requests like "I'd like 27 x86 servers for a class on distributed computing.. can you have it ready by next week?"
Blades fill a purpose, of which clusters are one.
-B
It's already been done, and done better than a stack of these little CD-sized guys. The RLX deals are pretty damn amazing. I've had occasion to see two different models in the past two years, and have been impressed each time. My favorite has to the be Transmeta-based blades, just because the consume like 9 watts when sitting idle. They're cool enough that you'd have a hard time telling they were powered on.
What makes something like an RLX chassis better than stacking in "little PCs" is that RLX has some very nice mgmt software that comes with the whole unit. Basically, you dedicate one blade to do mgmt stuff, and the rest (whether you have one chassis or ten) can all be managed by it. You can have all the blades sitting there blank, and remotely (and programmatically) boot up and then re-image any number of them with Windows or Linux, in any configuration you've set up. (The OS images are actually just tarballs of previously-installed operating systems you've set up and saved. So you can dedicate one blade to OS imaging duty, put Red Hat in whater config you want on it, upgrade the kernel or whatever and then push that tarball out to a "test blade" if you want to see how your apps runs.)
You also get more hardware with something like an RLX. The newer ones have dual fibre channel NICs, dual Gig Ethernet NICS, and a dedicated backplane network for "out of band" management, and an optional layer 2 switch for that chassis. That all means that you can make a cluster out of them really easily. And it means that you can do away with their hard drives, boot off the net and use network disk everywhere while still keeping them as "individual" servers. One more bonus: you don't have a cabling nightmare, and don't really need KVM for every server. They are also designed with heat output in mind. You can literally fill a 42U rack full of them (which is a total of like 330-something P3s) and still power it up. They're hot-swappable, too.
I don't work for RLX, I've just seen them up close a couple times (we're demoing one unit now, and will get another soon). If you are thinking of making a cheap cluster, or just want a lot of PCs in a little space withut a management headache, you might do well to look into RLX.
-B
-B
I think you're right. A a matter of fact, I think I can prove it.
Yep. I[ve proven that you are correct. Trouble is that Slashdot won't let me post my proof here, so I put it on my site.
-B
That depends on what damages I incurred, if any. If they offered to replace it before anything bad happened, then my estimation of Ford would go up quite a bit. But whatever my new love for Ford and their "doing the right thing" it still wouldn't change the fact that the original seatbelt was in some way sub-standard and needed to be replaced. Not a big deal, unless it was the latest in a long line of parts that needed replacing fr one reason or another.
Sheesh. I'm going to start Linux/Unix/FreeBSD/Sparc/OpenSourceSoftware bashing soon. Maybe that'll catch on.
Be my guest. You're allowed to hold whatever opinions you want. But you'd get more mileage out of constructive criticism rather than outright bashing, like I said eailer. You also wouldn't be the first to complain about the inadequacies of non-MS software. To paraphrase an old saying: All software sucks; some just sucks less.
Or maybe I'm just a dumbass like Boy Wonder down below claims. Ain't having choices grand?
-B
I don't know if anyone else had come up with a similar law before I thought of it a number of years ago (thanks mostly to the brilliant work of none other than Ivan Stang), so I'll put a flag in it right now and call it Wee's Law of Tinfoil Hats.
-B
You should if the seatbelt doesn't work. Hell, you might even be able to blame Ford is the seatbelts are too difficult to operate or hard to find. I once owned a car that wouldn't start unless the driver's seatbelt was latched. My newer cars don't have this "feature". Can I blame them for that?
Seriously, the MS bashing on this site is soooooooo lame.
Not really, no. It's incredibly easy, that I grant you. That makes for more bashing than necessary, much of it from an uninformed (or "partially formed") opinions. MS makes generally shoddy software, typically driven by a need for the cash generated by the upgrade mill rather than the needs of end users. Things like security are not a priority (or haven't been until recently) because of this, and they get a lot of flak for it. Rightly so, IMO. This makes it easy to come down hard on them.
Having said that, one should always try to use the right tool for the job. If that job is Grandpa's PC, then Windows is probably the right tool for it. If the job is my database server or my cluster, or even my mail server, Windows probably isn't. That's my opinion, and I'm completely allowed to call bullshit when I see people trying to use MS products for jobs for which they are ill-suited.
Put another way: using my personal experience/expertise to form an opinion (which may just come down harshly against MS) is never somthing I will think of as lame -- no matter what the venue. That line of reasoning leads one to suppose that MS is beyond the range of people's opinions, or somehow exempt from scrutiny. Such is not the case. (BTW, this is true of Linux, *BSD, Solaris, Oracle, et al. as well.)
If the opinions you see as bashing are obviously uninformed and useless, then call them as such, but don't label everything you don't particularly agree with as "lame MS bashing".
-B
Thanks to a couple ex-Loki guys (Sam Latinga and Mike Philips, most notably) this is often the case. I'm sorry to question your experience, but things weren't always this way. I recall waiting months before finally giving up on Linux T2 (I had purchased both Linux and Win32 copies).
Who's to say that the goodwill which keeps Linux patches coming will continue? If Sam goes on vacation for two weeks while a Win32 patch is released, what will Linux T2 players do? Play only on Linux servers until he gets back, that's what.
-B
I was a huge Tribes fan, and waiting for a long time for Tribes2 to come out since it meant that I could finally dump Windows and use Linux for everything. After about a year I realized that wanting to use Linux for everyday gaming use was a pipe dream. I still play Wolfenstein, SimCity, Tribes2 and old ROMs and stuff, but to play PC games these days you either need WineX or a Win32 partition.
-B
While I was indulging my data storage daydreams, I came across a discussion board thread which talks about the various issues surround storing digital media (pictures, in this case). It was pretty intersting reading. I hadn't thought about gold-plated CDs before, and that sounds like a great idea as long as the hardware to read them exists for the duration of the media's shelf life. Even NASA has been having trouble in that area.
At first blush, I'd say the way to save all the images would be some sort of distrubuted filesystem, a la Freenet. Package an ASCII metafile with the ROMs file format info along with the actual image file and that should do it. Some sort of centralized system of making sure that at least N copies exist in "the wild" and the data could be reasonably safe. I'm oversimplifying, of course, but it occurred to me that data integrity and file formats might not be the only barriers to long-term data storage. Governments aren't especially data-friendly 100% of the time, either. If you really want to save data for all posterity, you have to protect it from yourself as well.
-B
I can't see how it's even "quietly". Red Hat has had this info on their website for a while now (can't recall exactly when I saw it and submitted it to slashdot, but it was before last December). As for "microsoftish"? Well, you get what you pay for. Most of the apps I personally run are patchable from sources outside RH. At work, we tend to get patches from the "vendor": ssh from openssh.org, apache from httpd.apache.org, PHP from php.net, etc. in addition to ones from redhat.com/errata/. Likely, we'll shift that balance even more.
I guess Red Hat is being microsoftish by trying to make a profit (maybe someday), ... or maybe it's the windowsupdate.com like ability to patch over the web.
They already do make a profit. Not a large one, but they are profitable. Is this move intended to generate more revenue? You bet. I can imagine them saying "Well, if you want/need guaranteed errata, then buy Advanced Server. If all you need is Linux, download an ISO and then after a year patch yourself for free." That's fine.
This also certainly ties into their Red Hat Update service, which already does network patching, same as Windows Update. (Forgive me if you were making that point; I missed the meaning of "from the web", and am construing "web" to mean "network".)
I think they're more Microsoftish than you may think, and I say 'right on!'.
As long as I have source, and they don't force me to deal with choices they've made, I'm with you. As soon as Red Hat starts trying to remove my liberties by trying to think for me, like MS has made a mint doing, I'm switching. But dump them for trying to make money selling services to those who need/want them? What could be wrong with that? I'd rather download free ISO from a company that only supports it for one year than have to pay up PBS style.
-B
-B
At a previous company, I had four machines in my office, three of which had 21" monitors. They made me get rid of a couple monitors because they couldn't balance the AC. One of the environmental servcies guys told me that a single 21" monitor is the equivalent of 6 people, as far as cooling the room is concerned. I was pretty astounded by that.
-B
I admit that mplayer isn't the the most friendly install with the best UI, but if I cared enough to bitch about how it's laid out, I'd grab the source and effect some changes. It's easier to throw a fit than to fix things.
-B
I was mostly talking about the Dell workstations. They order 20 at a crack, all with flat panels, unless specifically ordered with a CRT. They also get them for server rooms. and whatnot.
Personally, I like my 21" Sony CRT at home much better than the 1800FP I'm using right now -- video, games, whatever, be damned. A CRT looks much better to me and is certainly easier on the eyes.
-B
I can tell you that those two things (well, one does, anyway) rank pretty high on a large organization's list. For example, I can tell you that any new computers which come with monitors bought by UCSD's CS dept have to be ordered with LCD monitors now. The power savings are pretty big, even though it may take a while to phase in the new machines and their flatpanels. A couple friends in various other large companies have also seen this trend.
My guess is that Sony is merely catering to business needs and pressures and not thinking of home users as much.
-B