Why? I'm still running RH7.3. The reason I didn't upgrade to 8.0 is mainly that the.0 releases are by many considered to be problematic. I tried 7.0 when it was released and I hated it. I had decided to upgrade from 7.3 to 8.1 as soon as 8.1 was released. Now I start wondering if I should rather stay on 7.3 and wait for 9.1 to be released. Or is it about time I try another distribution?
Yeah, according to RH's research, 80 of the RH users out there are running 7.x or higher. But i'd suspect not more than 30% or so are running redhat 8, and most of those are desktop's, i'm sure, not servers. Which brings up:
There's a much bigger problem: What about support for those of us running RH 7.x?? What happens when a vulnerability occurs in the code? How far back do you think they'll release patches? I know they don't continually patch 6.x now, so I'd imagine that they won't continue to patch 7.x after this release. That's going to leave me and my 50 computers at work running RH 7.x high and dry.
Cause, I'm damn sure not upgrading to RH 8. You may think it's buggy, but you don't know the half of it. Try running it on a server sometime - it CAN'T be done in a sane manner. The default install installed apache 2, but then tried to install a version of mod_perl that is incompatable with 2.0, so then it also installed 1.3.19, but then mod_php wouldn't work, no SSL support, etc. Good grief. RH 8 was buggy beyond belief.
So, now, i'm expected to update to something, either 8 or 9 on 50 comptuers at work, and not break anything in the upgrade?
Thanks, redhat. If you weren't what everyone asked for, i'd move back to debian or gentoo. This is exactly why no one wants to release binaries for linux. If you wrote something for NT 3.51, chances are it still works in 2k server. Not saying that IIS is better, but linux needs to work on the not forking so much thing, and leave some sanity in the backwards compatability.
Heh, not all of us can run the most popular pr0n site in the world =). While I commend your efforts, and am jealous of your lifestyle, seriously, my job is basically only slightly above zero stress.
Oh, and to the one beer a month at TGIF, it's more like "sharkey's" is two doors down, and has happy hour from ELEVEN to NINE every weekday, with $2 for 34oz domestic drafts, or 3.75 for 22oz highballs/hurricanes/whatever (long island iced tea in a 22oz size comes to mind).
Seriously, we come to work and hang out a lot. We play with some of the hardware, etc. We have fun.
I was just talking about this with my boss yesterday. As to getting tipsy on your lunch break: We don't so much condone it as we ignore it. And we don't so much ignore it as we celebrate it.
I'm 95% kidding, of course, but alcohol not being on the same list as comming to work on crack definately helps some. You're not afraid to order a draft beer or two with your sandwitch.
Plus, another big one is how comfortable the employees are in their work environment. We don't have an official uniform. I wear sandals and my Itchy and Scratchy t-shirt to work. BUT I feel more productive because I don't have to deal with wearing a company logo polo shirt and dress shoes.
A lot of smaller technology firms could benifit from expanding "dress down friday" to everyday. Really, how often do you see your customers? I know of one customer who currently lives in the same town as our office, and he signed up because he knew me. As long as we remain professional on the phone, what does it honestly matter?
Also: Music. We're allowed to quietly play music at work. Note: QUIETLY, because we have to be able to hear phone conversations, but, we all like different kinds of music (techno, country/classical, punk). Being able to have background music does help.
Plus, being able to browse websites not strictly related to work helps, too. We just call it "selective trolling" - keep the work URL in the.sig, and that gives us an excuse to read slashdot at work =). We explain to the owner that if we post to slashdot, people will see the link in the.sig file and sign up for service, so he's sort of alright with that.
In summary, what the hell is the RIAA worried about? I feel most people are like me, they pay for what they like, and try to do the honest thing.
Well, yeah, in a way. If i can get an album 128kbps mp3, I might go out and buy it later. But, on the other hand, if i can find it ~190 variable ogg, i don't bother.
I've seen very few portscans against any of my internet connected boxes.
What you should be looking for is the precursor to portscans: Broadcast storms.
We have some (12mhz) sparc IPC's, running about 16 rstatd graphs a piece, sitting on top of our book shelves at work, so that we can see the status of all our machines. A couple of times a day, we see the packet traffic spike on all machines simultaneously because of a packet storm, packets getting sent to the broadcast address, and all that.
After those determine what is and isn't a machine, then we start to see port scans.
should be the ability of playing online without any additional cost. IMO it should be left for the makers of the game to charge so they can keep making great games we all love to play.
I agree, however simple economics dictate otherwise. A one time purchase of a game for $50 (which, whatever, is somewhat reasonable) does not cover the ongoing cost of bandwidth and support. Same reason why people can't sell "lifetime of dial up internet for one price" or "lifetime webhosting, just pay $200 up front".
The longer the game is around, the further you go from recouperating the sunk costs, such that with infinite time, the game becomes infinately expensive for the company, yet it only has a fixed cost of $50 for the consumer.
There are only a couple of options here: the Everquest model (basically give the game away for free, but require subscription to play), the warcraft model (offer what I just claimed was impossible), or the FPS model (sell the game, let the loyal fans set up servers).
I honestly don't know how blizard doesn't charge for battlenet. I have no idea what lets them do that. They still support games of warcraft 2 and starcraft. WC2 came out IN 1995/6!!! They can't still be making money on it, which means anytime someone logs onto battle.net to play WC2, they lose money in bandwidth.
The problem is that it's a little hard to call Qwest back and say "hey, the form you sent us to fill out and mail back to you is an excel spreadsheet. Did you ever consider sending it as an RTF or a.txt file?
I'm guessing you mean faster processors. It doesn't take SMP to run openoffice. But hey, what do you expect from a troll, intelligence? bah!
No, he meant processors. Not as in multi procs for one system, but as in multiple machines. Obviously it's not newsworthy if Mass. is upgrading one machine. He meant multiple machines, processors with cost over many machines.
It's compatible with more than koffice. Word isn't compatible with much other than word. If Koffice is all you're using, why the fuck do you need it to support word?
Because word is the world standard for written documents in the professional business world. Hate to break it to you, but where I work, we don't have but 2 windows machines, and one running VMware, out of about 80 computers. We get lease documents, legal notices, business proposals, ad nauseum, in word or excel format. If you can't read it, you limit your professional image and connectivity. K-office is compatable with k-office. Open/Star office at least has basic word compatability and functionality. Please, microsoft may suck for their draconian EULA's, their extremely high prices, their business model, etc. But they make a good office suite. Plus, like it or not, it's the world standard.
I mean, if the computers were built for a specific purpose, and they're still used for that purpose, why upgrade?
Reasons to upgrade: 1.) Your programs require more system resources. This is fair. We were using QuickBooks from ages ago until they stopped providing tax tables for our version, forcing us to upgrade *grr* and the new version has new bells and whistles so that it bogs down the P-90 w/ 32 megs of ram. 2.) You want support from Microsoft. But, then, if you really wanted to install all the updates for windows 95, wow. That's a lot of updates, probably adding enough to your system to bog it down alone.
But, then, why not upgrade the hardware and install the same copies of Win95? You'd be surprised how many programs will work with win95.
Or, how much do new copies of windows 98 cost? I don't know if they're still available, or how that works. You may have to do the MS stupid "upgrade to downgrade" thing.
If you want to keep windows there are lots of alternatives to look at. I say this because developing new software for linux and training your average high school grad 40 year old secretary to use linux won't be cheap. Something like RedHat 8 is intuitive, but it ain't perfect. Keep in mind that intuitive doesn't mean everything - familiarity is much more important.
I'm all for linux, but I'm also all for lowering the TCO. And i know that over time, linux is definately cheaper. But, then, how many politicians look long term? You look short term so that you get re-elected. Long term politicians get voted out of office.
He estimated that an internship at Microsoft may pay as much as $25 an hour, or $1,000 a week.
I can confirm this. I had a friend who worked as a MS intern over the summer. He got his CS degree from Va. Tech, and then moved to redmond. I asked him what he was working on, and he said that he didn't want to get into specifics, but it was mostly windows CE. I asked him how much he would be making to start out. He said he couldn't say. I asked, "well, I'd guess about $50k/yr". And he shot mt. dew out his nose and said he had been making that as an intern.
Dave Wrighton, are you out there? IM me or tess, let us know how you're doing.
Oh, he also said that a lot of the higher up MS employees basically have money pissing contests, i.e. build a $600,000 authentic italian pizza oven in your backyard, etc.
Programmers, Editors, OC3s and Racks of web servers cost money.
Don't lie. You don't have racks of webservers!
As I'm referencing the FAQ, I see: 5 load balanced Web servers dedicated to pages 3 load balanced Web servers dedicated to images 1 SQL server 1 NFS Server
The 8 webservers are described as: PIII/600 MHz 512K cache 1 GB RAM 9.1GB LVD SCSI with hot swap backplane Intel EtherExpress Pro (built-in on moboard) Intel EtherExpress 100 adapter
Now, I know all of those could be 1U very easily (only one hard drive and one PCI card), but let's say they're 2U's.
The NFS server is described as: Dual PIII/600 MHz 2 GB RAM (2) 9.1GB LVD SCSI with hot swap backplane Intel EtherExpress Pro (built-in on motherboard) Intel EtherExpress 100 adapter
Again, this could be a 1U, but let's say, since it's a Dual system, it's a 2U. I know for a fact, this could all fit with acres to spare in a 2U (we have 3 of them on our network with dual PIII-1.4's, 4GB ram, and 6x73GB SCSI drives). In fact, just for the sake of arguement, let's call this a 4U.
Now, the SQL monster server: Quad Xeon 550 MHz, 1MB cache 2 GB RAM 6 LVD disks, 10000 RPM (1 system disk, 5 disks for RAID5) Mylex Extreme RAID controller 16 MB cache Intel EtherExpress Pro (built-in on motherboard) Intel EtherExpress 100 adapter
In this one, you've not only got Quad procs, but you have 2 full size PCI cards you have to deal with, as well as you have to find somewhere to put 6 hard drives. We'll call this one a huge, massive monster at 8U's.
After all that, we have to add the Cisco equipment: [quote] All boxes are networked together through a Cisco 6509 with 2 MSFCs and a Cisco 3500 so we can rearrange our internal network topology just by reconfiguring the switch. Internet connectivity to/from the outside world all flows through an Arrowpoint CS-800 switch which acts as both a firewall load balancer for the front end Web servers. [/quote] I don't know how big these cisco's are, but let's say these 2 cisco pieces and the arrowpoint are 10U's (say, mabey 4 for each cisco and 2 for the arrowpoint). I see this as very reasonable.
And now the tally!
16U for 8 webservers 04U for 1 NFS server 08U for 1 SQL server 10U for equipment -------------- 38U total.
Most racks are 42 U's. With this, you even have space for a 3U APC battery backup and a 1U power octopus. So, unless you're just keeping your single proc, single hard drive systems in 8U servers, and putting your leftover pizza in there to keep it warm for lunch, you're wasting space!
They DON'T want to do mirrors - a couple of reasons. Slashdot hosted mirrors: Bandwidth != free. Slashdot supported user hosted mirrors: legal and/or statistical reasons (banner ad displays, click thru's, page views, etc.)
Mirrors probably aren't going to happen on an official level, folks. Just keep posting them in the stories, like ya do now.
I'm guessing slashdot probably gets more submissions than Fark. I'd also be surprised if lots of those submissions weren't duplicates.
Fark solves this very simply. You MUST submit a URL that corresponds with your story. Any URL can only be submitted once. Therefore, you can't submit cnn.com, www.wwe.com, etc, because chances are some dude already has. This encourages you to only submit new news articles, and ones with a format like http://www.newssite.com/stories/top/2003/march/19/ 3.html
Yeah, the signal to noise ratio on total fark is higher than fark, and yeah, it would probably be higher still on slashdot, but it would certainly get rid of a lot of dupes if they did this.
Hehe. Dude, I'm in college, working on my history degree. I'm taking an in depth study of the fate of the roman empire, called "later roman empire", we're going from rome to byzantium, from constantine the great (312) to constantine the 14th (1453).
heh, I got on AOL back when they put out mac versions before PC versions. When you had to download a butt-load of stuff to be able to use mosaic. I do have some sort of nostalgia.
I also used the same screen name, well for a while at least. I've had the screen name I use now since 1996 (which is redneck669), and before that I had redneck200 since about 1993(ish).
I don't know what the X means, but the P stands for "Piece Of Crap"
I assume you're talking about "XP".
XP stands for Jesus Christ. When the Emperor Constantine fought for control of the western roman empire at the Milvian bridge in 312, he supposedly saw the sign "Chi-Rho" (Greek Letters X and P) in the sky, along with hearing a voice which said "in this sign, you will conqueror". Chi-Rho, the way it is usually depicted in ancient artwork, is an X super-imposed on a P. Chi and Rho are the first two letters of the Greek name for Christ, pronounced "Kreestos". Hence, where we get "X-mas". I once heard a baptist preacher say that x-mas was bad because they were crossing out christ, x-ing him out. This is stupid - since the 500's X has been a sign for Christ.
Hence, WindowsXP is really Windows, version christ.
*obligitory comments* 1.) bluetooth is dead 2.) bluetooth is the future.
All that aside, one of the absolutely coolest things is the bluetooth enabled headset that goes with it. That is just cooler than anything i've ever seen.
Next project: make a NES emulator for it (hey, it's got 2MB rom!)
there's no provision for determining whether the server is down or whether it's just taking time to respond.
We had a problem with this once - Someone had misconfigured our main server (nfs server), or missed the memo on which servers were to be standalone, etc.
The NIS (password) server is STANDALONE. It boots all by it's self. For some reason, someone had NFS mounted something asenine (/adm,/var/mail, something) onto the NIS server.
Then, one day, we tested a UPS, and had a catastrophic failure (batteries not working, insides sparking, rebooting an entire server rack). The NIS server and the NFS server both went down.
Well, the NIS server wouldn't come up because it would hang trying to mount the NFS mount, and the NFS server wouldn't come up because solaris is set up by default to connect to the NIS server before it starts the NFS server. When it couldn't find the NIS domain, we knew we were screwed. So we had to scramble to find our "CD on an external scsi hard disk" to boot off of to change the mount directive in the NIS server.
Needless to say, everyone got the memo.
But, yeah, if you're mounting things around the network like we are, NFS is the ONLY way to go, and solaris implements it beautifully, except for the unexplained hanging, the non-verbose error messages, and the startup defaults, but, hey! who's counting? I've still had less problems with solaris NFS than Linux. We use NFS to mount/adm (our home directories and stuff) on most of our machines, we mount/var/mail on all shared hosting machines (two mail servers which aren't the hosting servers), we mount/opt/bin,/opt/sbin, etc on some of our clients machines to save them hard disk space, etc.
Slashdot Mirror
Why? I'm still running RH7.3. The reason I didn't upgrade to 8.0 is mainly that the .0 releases are by many considered to be problematic. I tried 7.0 when it was released and I hated it. I had decided to upgrade from 7.3 to 8.1 as soon as 8.1 was released. Now I start wondering if I should rather stay on 7.3 and wait for 9.1 to be released. Or is it about time I try another distribution?
Yeah, according to RH's research, 80 of the RH users out there are running 7.x or higher. But i'd suspect not more than 30% or so are running redhat 8, and most of those are desktop's, i'm sure, not servers. Which brings up:
There's a much bigger problem:
What about support for those of us running RH 7.x?? What happens when a vulnerability occurs in the code? How far back do you think they'll release patches? I know they don't continually patch 6.x now, so I'd imagine that they won't continue to patch 7.x after this release. That's going to leave me and my 50 computers at work running RH 7.x high and dry.
Cause, I'm damn sure not upgrading to RH 8. You may think it's buggy, but you don't know the half of it. Try running it on a server sometime - it CAN'T be done in a sane manner. The default install installed apache 2, but then tried to install a version of mod_perl that is incompatable with 2.0, so then it also installed 1.3.19, but then mod_php wouldn't work, no SSL support, etc. Good grief. RH 8 was buggy beyond belief.
So, now, i'm expected to update to something, either 8 or 9 on 50 comptuers at work, and not break anything in the upgrade?
Thanks, redhat. If you weren't what everyone asked for, i'd move back to debian or gentoo. This is exactly why no one wants to release binaries for linux. If you wrote something for NT 3.51, chances are it still works in 2k server. Not saying that IIS is better, but linux needs to work on the not forking so much thing, and leave some sanity in the backwards compatability.
Heh, not all of us can run the most popular pr0n site in the world =). While I commend your efforts, and am jealous of your lifestyle, seriously, my job is basically only slightly above zero stress.
Oh, and to the one beer a month at TGIF, it's more like "sharkey's" is two doors down, and has happy hour from ELEVEN to NINE every weekday, with $2 for 34oz domestic drafts, or 3.75 for 22oz highballs/hurricanes/whatever (long island iced tea in a 22oz size comes to mind).
Seriously, we come to work and hang out a lot. We play with some of the hardware, etc. We have fun.
I was just talking about this with my boss yesterday. As to getting tipsy on your lunch break: We don't so much condone it as we ignore it. And we don't so much ignore it as we celebrate it.
.sig, and that gives us an excuse to read slashdot at work =). We explain to the owner that if we post to slashdot, people will see the link in the .sig file and sign up for service, so he's sort of alright with that.
I'm 95% kidding, of course, but alcohol not being on the same list as comming to work on crack definately helps some. You're not afraid to order a draft beer or two with your sandwitch.
Plus, another big one is how comfortable the employees are in their work environment. We don't have an official uniform. I wear sandals and my Itchy and Scratchy t-shirt to work. BUT I feel more productive because I don't have to deal with wearing a company logo polo shirt and dress shoes.
A lot of smaller technology firms could benifit from expanding "dress down friday" to everyday. Really, how often do you see your customers? I know of one customer who currently lives in the same town as our office, and he signed up because he knew me. As long as we remain professional on the phone, what does it honestly matter?
Also: Music. We're allowed to quietly play music at work. Note: QUIETLY, because we have to be able to hear phone conversations, but, we all like different kinds of music (techno, country/classical, punk). Being able to have background music does help.
Plus, being able to browse websites not strictly related to work helps, too. We just call it "selective trolling" - keep the work URL in the
In summary, what the hell is the RIAA worried about? I feel most people are like me, they pay for what they like, and try to do the honest thing.
Well, yeah, in a way. If i can get an album 128kbps mp3, I might go out and buy it later. But, on the other hand, if i can find it ~190 variable ogg, i don't bother.
I've seen very few portscans against any of my internet connected boxes.
What you should be looking for is the precursor to portscans: Broadcast storms.
We have some (12mhz) sparc IPC's, running about 16 rstatd graphs a piece, sitting on top of our book shelves at work, so that we can see the status of all our machines. A couple of times a day, we see the packet traffic spike on all machines simultaneously because of a packet storm, packets getting sent to the broadcast address, and all that.
After those determine what is and isn't a machine, then we start to see port scans.
Yes. According to the storyline, shane mcmahon bought WCW, and then WCW invaded the WWF.
Lots of current stars (BookerT, Jamie Noble, etc) are from WCW.
expand the demographic of the show, and make it match their existing audience (WWF fans as you point out).
Not to split hairs, but TNT's audience would have been WCW fans, not WWF. WCW-Monday Nitro was on TNT, currently Monday Night WWE-Raw is on TNN.
Or, just use the virt user table in sendmail. Create a new alias everytime you give out an address.
should be the ability of playing online without any additional cost. IMO it should be left for the makers of the game to charge so they can keep making great games we all love to play.
I agree, however simple economics dictate otherwise. A one time purchase of a game for $50 (which, whatever, is somewhat reasonable) does not cover the ongoing cost of bandwidth and support. Same reason why people can't sell "lifetime of dial up internet for one price" or "lifetime webhosting, just pay $200 up front".
The longer the game is around, the further you go from recouperating the sunk costs, such that with infinite time, the game becomes infinately expensive for the company, yet it only has a fixed cost of $50 for the consumer.
There are only a couple of options here: the Everquest model (basically give the game away for free, but require subscription to play), the warcraft model (offer what I just claimed was impossible), or the FPS model (sell the game, let the loyal fans set up servers).
I honestly don't know how blizard doesn't charge for battlenet. I have no idea what lets them do that. They still support games of warcraft 2 and starcraft. WC2 came out IN 1995/6!!! They can't still be making money on it, which means anytime someone logs onto battle.net to play WC2, they lose money in bandwidth.
The problem is that it's a little hard to call Qwest back and say "hey, the form you sent us to fill out and mail back to you is an excel spreadsheet. Did you ever consider sending it as an RTF or a .txt file?
I'm guessing you mean faster processors. It doesn't take SMP to run openoffice. But hey, what do you expect from a troll, intelligence? bah!
No, he meant processors. Not as in multi procs for one system, but as in multiple machines. Obviously it's not newsworthy if Mass. is upgrading one machine. He meant multiple machines, processors with cost over many machines.
It's compatible with more than koffice. Word isn't compatible with much other than word. If Koffice is all you're using, why the fuck do you need it to support word?
Because word is the world standard for written documents in the professional business world. Hate to break it to you, but where I work, we don't have but 2 windows machines, and one running VMware, out of about 80 computers. We get lease documents, legal notices, business proposals, ad nauseum, in word or excel format. If you can't read it, you limit your professional image and connectivity.
K-office is compatable with k-office. Open/Star office at least has basic word compatability and functionality.
Please, microsoft may suck for their draconian EULA's, their extremely high prices, their business model, etc. But they make a good office suite. Plus, like it or not, it's the world standard.
Touche, troll. Touche.
~Will
Not to ask the obvious, but why upgrade?
I mean, if the computers were built for a specific purpose, and they're still used for that purpose, why upgrade?
Reasons to upgrade:
1.) Your programs require more system resources. This is fair. We were using QuickBooks from ages ago until they stopped providing tax tables for our version, forcing us to upgrade *grr* and the new version has new bells and whistles so that it bogs down the P-90 w/ 32 megs of ram.
2.) You want support from Microsoft. But, then, if you really wanted to install all the updates for windows 95, wow. That's a lot of updates, probably adding enough to your system to bog it down alone.
But, then, why not upgrade the hardware and install the same copies of Win95? You'd be surprised how many programs will work with win95.
Or, how much do new copies of windows 98 cost? I don't know if they're still available, or how that works. You may have to do the MS stupid "upgrade to downgrade" thing.
If you want to keep windows there are lots of alternatives to look at. I say this because developing new software for linux and training your average high school grad 40 year old secretary to use linux won't be cheap. Something like RedHat 8 is intuitive, but it ain't perfect. Keep in mind that intuitive doesn't mean everything - familiarity is much more important.
I'm all for linux, but I'm also all for lowering the TCO. And i know that over time, linux is definately cheaper. But, then, how many politicians look long term? You look short term so that you get re-elected. Long term politicians get voted out of office.
~Will
He estimated that an internship at Microsoft may pay as much as $25 an hour, or $1,000 a week.
I can confirm this. I had a friend who worked as a MS intern over the summer. He got his CS degree from Va. Tech, and then moved to redmond. I asked him what he was working on, and he said that he didn't want to get into specifics, but it was mostly windows CE. I asked him how much he would be making to start out. He said he couldn't say. I asked, "well, I'd guess about $50k/yr". And he shot mt. dew out his nose and said he had been making that as an intern.
Dave Wrighton, are you out there? IM me or tess, let us know how you're doing.
Oh, he also said that a lot of the higher up MS employees basically have money pissing contests, i.e. build a $600,000 authentic italian pizza oven in your backyard, etc.
~Will
Yeah, but my car doesn't stop working in 30 days if it doesn't get registered with the manufacturer.
Plus, if I want to, I can sell my perfectly working car to another consumer, in order to recoup my losses and buy another car.
Or, I can add performance headers, do a hybrid H22A/b16A6 engine swap, add a turbo.
Sweet, dude!
I was just poking fun, anyway. More computers, in my opinion, is a good thing.
Programmers, Editors, OC3s and Racks of web servers cost money.
Don't lie. You don't have racks of webservers!
As I'm referencing the FAQ, I see:
5 load balanced Web servers dedicated to pages
3 load balanced Web servers dedicated to images
1 SQL server
1 NFS Server
The 8 webservers are described as:
PIII/600 MHz 512K cache
1 GB RAM
9.1GB LVD SCSI with hot swap backplane
Intel EtherExpress Pro (built-in on moboard)
Intel EtherExpress 100 adapter
Now, I know all of those could be 1U very easily (only one hard drive and one PCI card), but let's say they're 2U's.
The NFS server is described as:
Dual PIII/600 MHz
2 GB RAM
(2) 9.1GB LVD SCSI with hot swap backplane
Intel EtherExpress Pro (built-in on motherboard)
Intel EtherExpress 100 adapter
Again, this could be a 1U, but let's say, since it's a Dual system, it's a 2U. I know for a fact, this could all fit with acres to spare in a 2U (we have 3 of them on our network with dual PIII-1.4's, 4GB ram, and 6x73GB SCSI drives).
In fact, just for the sake of arguement, let's call this a 4U.
Now, the SQL monster server:
Quad Xeon 550 MHz, 1MB cache
2 GB RAM
6 LVD disks, 10000 RPM (1 system disk, 5 disks for RAID5)
Mylex Extreme RAID controller 16 MB cache
Intel EtherExpress Pro (built-in on motherboard)
Intel EtherExpress 100 adapter
In this one, you've not only got Quad procs, but you have 2 full size PCI cards you have to deal with, as well as you have to find somewhere to put 6 hard drives. We'll call this one a huge, massive monster at 8U's.
After all that, we have to add the Cisco equipment:
[quote] All boxes are networked together through a Cisco 6509 with 2 MSFCs and a Cisco 3500 so we can rearrange our internal network topology just by reconfiguring the switch. Internet connectivity to/from the outside world all flows through an Arrowpoint CS-800 switch which acts as both a firewall load balancer for the front end Web servers. [/quote]
I don't know how big these cisco's are, but let's say these 2 cisco pieces and the arrowpoint are 10U's (say, mabey 4 for each cisco and 2 for the arrowpoint). I see this as very reasonable.
And now the tally!
16U for 8 webservers
04U for 1 NFS server
08U for 1 SQL server
10U for equipment
--------------
38U total.
Most racks are 42 U's. With this, you even have space for a 3U APC battery backup and a 1U power octopus. So, unless you're just keeping your single proc, single hard drive systems in 8U servers, and putting your leftover pizza in there to keep it warm for lunch, you're wasting space!
You slashdot editors: you're always braggin!
~Will
it might be good to keep a mirror link list / submission form for all users, even once it's posted
As slashdot has covered this NUMEROUS times (obviously, as it's in their FAQ).
They DON'T want to do mirrors - a couple of reasons.
Slashdot hosted mirrors: Bandwidth != free.
Slashdot supported user hosted mirrors: legal and/or statistical reasons (banner ad displays, click thru's, page views, etc.)
Mirrors probably aren't going to happen on an official level, folks. Just keep posting them in the stories, like ya do now.
I'm guessing slashdot probably gets more submissions than Fark. I'd also be surprised if lots of those submissions weren't duplicates.
/ 3.html
Fark solves this very simply. You MUST submit a URL that corresponds with your story. Any URL can only be submitted once. Therefore, you can't submit cnn.com, www.wwe.com, etc, because chances are some dude already has. This encourages you to only submit new news articles, and ones with a format like http://www.newssite.com/stories/top/2003/march/19
Yeah, the signal to noise ratio on total fark is higher than fark, and yeah, it would probably be higher still on slashdot, but it would certainly get rid of a lot of dupes if they did this.
I'm going to CC this to CmdrTaco.
Purple title bars! We want PURPLE title bars!
You want it, you got it!
Don't you mean "jesus christ, he's right?".
Hehe. Dude, I'm in college, working on my history degree. I'm taking an in depth study of the fate of the roman empire, called "later roman empire", we're going from rome to byzantium, from constantine the great (312) to constantine the 14th (1453).
I know things.
heh, I got on AOL back when they put out mac versions before PC versions. When you had to download a butt-load of stuff to be able to use mosaic. I do have some sort of nostalgia.
I also used the same screen name, well for a while at least. I've had the screen name I use now since 1996 (which is redneck669), and before that I had redneck200 since about 1993(ish).
I don't know what the X means, but the P stands for "Piece Of Crap"
I assume you're talking about "XP".
XP stands for Jesus Christ. When the Emperor Constantine fought for control of the western roman empire at the Milvian bridge in 312, he supposedly saw the sign "Chi-Rho" (Greek Letters X and P) in the sky, along with hearing a voice which said "in this sign, you will conqueror". Chi-Rho, the way it is usually depicted in ancient artwork, is an X super-imposed on a P. Chi and Rho are the first two letters of the Greek name for Christ, pronounced "Kreestos".
Hence, where we get "X-mas". I once heard a baptist preacher say that x-mas was bad because they were crossing out christ, x-ing him out. This is stupid - since the 500's X has been a sign for Christ.
Hence, WindowsXP is really Windows, version christ.
But, of course, we're AOL and this is Slashdot, so naturally everything we do is wrong
You got me on the internet.
Granted, I've since graduated, but *blush* you were my first.
*obligitory comments*
1.) bluetooth is dead
2.) bluetooth is the future.
All that aside, one of the absolutely coolest things is the bluetooth enabled headset that goes with it. That is just cooler than anything i've ever seen.
Next project: make a NES emulator for it (hey, it's got 2MB rom!)
there's no provision for determining whether the server is down or whether it's just taking time to respond.
/var/mail, something) onto the NIS server.
/adm (our home directories and stuff) on most of our machines, we mount /var/mail on all shared hosting machines (two mail servers which aren't the hosting servers), we mount /opt/bin, /opt/sbin, etc on some of our clients machines to save them hard disk space, etc.
We had a problem with this once - Someone had misconfigured our main server (nfs server), or missed the memo on which servers were to be standalone, etc.
The NIS (password) server is STANDALONE. It boots all by it's self. For some reason, someone had NFS mounted something asenine (/adm,
Then, one day, we tested a UPS, and had a catastrophic failure (batteries not working, insides sparking, rebooting an entire server rack). The NIS server and the NFS server both went down.
Well, the NIS server wouldn't come up because it would hang trying to mount the NFS mount, and the NFS server wouldn't come up because solaris is set up by default to connect to the NIS server before it starts the NFS server. When it couldn't find the NIS domain, we knew we were screwed. So we had to scramble to find our "CD on an external scsi hard disk" to boot off of to change the mount directive in the NIS server.
Needless to say, everyone got the memo.
But, yeah, if you're mounting things around the network like we are, NFS is the ONLY way to go, and solaris implements it beautifully, except for the unexplained hanging, the non-verbose error messages, and the startup defaults, but, hey! who's counting? I've still had less problems with solaris NFS than Linux.
We use NFS to mount
NFS rox.