plus explaining to 130,000 kids how to rpm -Uvh rpmname.rpm isn't optimal
There are plenty of automated software management systems out there, commercial and free. At the very least, create a tiered package deployment infrastructure. Setup the laptops with a system based on RH9 and cron-apt package. Then central IT in Lansing tests and approves new RPMs to be pushed out to the students. They serve up the new RPMs on a centralized server, and each school district mirrors the site. Each laptop uses cron-apt to update itself from the local mirror during homeroom. You don't even have to consider running the Ximian updater, just apt for RPM and good ole cron.
Each laptop runs sshd and has a HelpMe! application which really just opens an IM window to a Jabber chat room hosted by Central IT. Worst comes to worst, someone from the central help desk ssh's into the kid's laptop to fix the problem, or the kid drags the dead laptop to the school's tech liasion. A central help desk of 5 or 10 and a lead sysadmin for 130,000 kids doesn't sound too expensive to me. Sounds pretty damn beautiful in fact. Remember, if an app (like a jabber client) does something you don't want it to do (like turn each laptop into a chat whore), you have the source and can actually do something about it.
If they have even a little money to spend they can use IBM Director or even CA's AMO and SDO (cough) for commerical inventory and software delivery products.
No sixth graders having to learn RPM, sorry. Nor any sixth graders getting their laptops infected by opening their email or visiting a malware site. I had a chance to drive the development of Linux laptops for 100,000 students nationwide. Too bad the organization ran out of money doing all the other stuff you need to do to run a school system.
HBO On Demand just started on Oct-1 and I can play any show or movie airing during the month at any time without incurring extra charges. Just like the commercial says, you can play, pause, fast forward and rewind any On Demand show right from the remote. Add that to my 144-hour TiVo, and its a crankin' setup. And those movies are separate from the virtual Blockbuster in my cable box with hundreds of titles for $3.99 and a 24-hour viewing window. So your nirvana may not be at hand, but having 500 movies and shows at my fingertips (and 1/3rd at no extra charge) is a pretty good start.
Each time this conversation comes up, I'm surprised no one mentions the Atom Films On Demand channel carried by Comcast. There's some cool stuff in there.
stream you the video of the concert to you[sic] PDA
I could have used something like at the performance of Chicago I saw recently. We sat so far over to one side, I had to duck down to peer below a stage light just to see the right half of the stage. But we had the privilege of paying the same price as the folks who sat front row center, so I was grateful for that.
If anyone does it, the guys at LUFS will write a driver for exactly that. "Remember, everything is a file, and if it isn't, it should be!"
I can see a driver that posts to encrypted anonymous groups optionally via anonymous remailers and then checks google for updates. Post 2 messages per chunk: a pointer to each chunk's subject line, and each chunk. The data in the pointer-posts should be enough to formulate a directory structure. (Hey, I might be onto something here...) Now, for the UI. Something that can generate posts and reconstruct files. Perhaps an ftp-like interface? Or just a CLI that can be backgrounded for the reconstruction phase.
It seems like that would give you much better access to vast data storage without having to staple together slow, medium, and fast-access data. If they do write drivers for all those methods, I vote for this to be the first one they tackle:)
I look at it this way.. you can "cron" up Windows Update or apt-get update; apt-get upgrade. However, Windows boxes will typically have to be rebooted and if you're running 24x7 you better have them clustered (with different auto-update times). And that goes wayy beyond all the stuff about trusting and automatically installing Windows patches.
At least on the Linux side, there's a little hope. Suppose you have 3 tiers.. you can setup a box that will keep debs or rpms updated via cron. apt-get update; apt-get upgrade -d IIRC. Then in the next tier, you have a sample of the different types of machines that can be found in your organization -- web, db, file, directory, whatever. Configure those boxes to point to a web site on your 1st level box in their apt sources. You can manually copy updated debs or rpms into the 'testing' website and let the beta boxes automatically update themselves using cron-apt. Test them all, check to make sure functionality is unaffected.
Then, set the last tier boxes (your production level boxes) to use a different virtual web server (say, "production") on the 1st tier box as their apt sources. Once you have tested and approved the patches applied to the 2nd tier boxes, you can copy the new debs or rpms to the DocumentRoot that serves the 3rd tier boxes. Et voila! Staged releases just by copying a few hundred k to a couple of directories on one box, spaced out over the course of a week. No more "patch money [sic]".
See cron-apt for reference. And yes, people, RH has apt in all its dependency-checking glory. So move along.
Back in January, I was staying at the Hilton Times Square, and while hurrying back inside one night, I looked up at all the monitors over top a neighboring shop, probably Madame Tussaud's Wax Museum. Sure enough, three of the monitors were displaying pop-up spam right over top of their regular advertising!
Who could imagine having that much tech and throwing it all away by directly attaching Windows boxes to the Internet without a firewall. Have they no shame!? Its been several months since I've looked, but I'm sure someone has had their way with those machines by now. Now THAT would be a fun box to 0wn. Pair that up with a laptop on wireless and you can have some serious fun with passers-by.
Sounds like I have a much smaller circle of friends. Remember, too, that when you put up your own Jabber server, you can still connect with your contacts on those privately owned systems. The difference is that if they prefer to use a "chat aggregation client" like Gaim or Trillian, you will be able to offer them an alternative when the owners of their preferred IM system throw the big red switch on 3rd party clients.
Jabber was not built to just switch everyone over to a new, privately owned network. It was designed from the beginning to allow groups, defined any way you want, to run their own server, with its own custom features, but still link to any outside networks with their own set of users, either Jabber, AIM, Yahoo!, MSN, or anything else.
Finally pushed me to install a Jabber server for my circle of friends. I've been using Gaim for a year or so, and its time to get my wife off Trillian ASAFP. I'm just sick of using closed source apps against a few servers who's owners occasionally show a propensity to piss all of us off.
Nice to have an alternative available that *I* control, end to end. And no, the Gaim Win32 interface does not suck. Just use the Wimp theme.
Exactly. I helped build NYU CompSci's very first web site and spent many days converting the technical paper collection to PS when electronically available and scanned to TIFF when it wasn't.. like for papers dating back to the late 60's.
I guess I *am* new around here, because I figured more people here would understand that the b2b uses of the 'net will make all the consumer-level stuff seem like a speck, a blip on the radar, miniscule and insignificant. A million dollars will be spent linking enterprises for every thousand dollars spent by consumers.
If you're only looking at web pages, you're missing 75% of the traffic that traverses the Internet today.
It doesn't need any of those privileges, but Linux has no mechanism to protect you on that level
Obligatory reference to LIDS. LIDS lets you specify exactly which users and applications have any combination of an extensive list of privileges, including read or write privileges at the file level and opening sockets. A common example would be hiding/etc/shadow from everyone (even root) *except*/bin/login, who has read-only access.
My old quote used to be "can't get root if there is no root", but you weren't claiming that Linux suffers from multiple privilege escalation vulnerabilities. All told, you are right about native Linux, but there is at least one fix.
Re:I'm as stumped as my girlfriend usually is
on
Telstar 4 is Down
·
· Score: 1
All told, I think "CW beacon" would have been the best choice. Many/most/all(?) satellites do transmit a CW identifier for a number of purposes, including the one the author used it for. He listened for the "I am alive" morse code transmission from the sat and heard it. So the sat didn't explode or unexpectedly fire its repositioning jets and move out of its assigned orbital slot.
Re:I'm as stumped as my girlfriend usually is
on
Telstar 4 is Down
·
· Score: 1
A 12 year old kid will NOT be able to compete in tomorrow's marketplace or the global community if she doesn't grow up Internet-connected today. If that mother disconnects the cablemodem, one more kid from the next generation will be lost.
But satellite tv has no redeeming educational value, and kids running around half naked eating dirt is just funny as hell.
perfect /. timing yet again (GRUB help?)
on
Booting Linux Faster
·
· Score: 0, Offtopic
I'm trying to get a quirky PC to boot Debian 3r1, and I can't get it to boot without the rescue floppy.
LILO gave L 40 40 (repeat indefinitely) at first. I added
device hda=0x80
to lilo.conf and that replaced the L 40's with just LI. Compiled a 2.4 kernel, and now I get
LILO L
instead of
LILO Uncompressing Linux......................
so I decided to try GRUB, and now the system boots to the grub CLI, and when I try to specify my 2.4 kernel, I get Error 13, unknown or unsupported kernel format. I'm currently recompiling the kernel as a.out format instead of ELF, but it'll take another hour.
Anyone have any tips? And no, the PC isn't bad, or the RAM, or the hdd's.. I've had this machine running for about four years.. it ran multiple versions of RH6 and 7 before they decided to require a minimum of 64MB RAM. Any help on getting GRUB working would be appreciated.
And if you stick around long enough at DeveloperWorks, you'll make a grid out of those PCs and offer the district more compute power than they ever realized they had.
I'd be more concerned with an attempt to take one out on takeoff than landing
That's what happened in Kenya several months ago. A SAM-7 was fired at a chartered Israeli jet that had taken off a few seconds earlier. The terrorist fired it off too early, and the SAM never picked up on the IR signature you worry about. The pilots felt a bump on the left side of the plane and saw the smoke trail as the rocket flew past the plane. Once they reached Israeli airspace, they were escorted in by Israeli F-15's and landed safely.
I'm surprised you don't seem to remember this event. Its exactly what you describe. Except, of course, the missile failed to destroy the target.. more like bounced off it.
I, like many here I would guess, have the BEFW11S4 (i think) wireless AP/router/4-port switch model. One thing that I would change right away is to make a "real" DMZ. I'm sure most people here know enough to not put a box in the "DMZ" which really just means "no firewall for this one box", but instead forward ports in to a specific box.
I would want to be able to firewall off an Internet-accessible box from the rest of my network while forwarding one or more inbound ports to that box. That would be better than all the port shenanigans you have to go through to support NetMeeting, for instance.
As a tech no-coder for ~10 years, I don't think HF is the critical band. Not to disrespect the very foundation of our hobby, but, IIRC, we have 10.0GHz to 10.5GHz and DirecTV has 10.5GHz to 11.0GHz. I believe we have the exact same bandwidth as them, and butt right up against them (IOW identical RF properties). I think its fair to say that they do a wee bit more with their allocation (which cost mega $$$) than we do with ours.
Back in y2k, the ARRL put up a website asking us for ideas as to where ham radio should go from here. They wanted totally new, uheard of ideas that may spark some interest by the manufacturers. I posted a warning that we need to focus on the 10GHz band so that we don't risk losing it. Remember HF is 30MHz wide, and we get, what 3MHz of it all told? We have spots in 1.2Ghz, 3GHz, 5GHz, 10GHz, 24GHz, and 46GHz and we have **500MHz** in the 10GHz band alone.
The technology I proposed was one that focused on VOIP using high speed data links in cells and linking cells together using even higher speed links. I suggested using a class B multicast subnet to "channelize" frequencies. Then, as clubs upgraded their cell-linking technologies, they could extend a "channel's" reach and offer additional channels to their local users. As an example, the most popular repeater frequencies in NYC could be extended all the way out to Rochester by using extenders in the 10GHz range, while clubs along the path could choose from the selection of channels they wish to carry. I specifically suggested keeping a separate head unit to allow individual users to upgrade the link technology and gradually utilize more channels at their location.
Sure enough, in the latest CQ VHF, I saw a product coming soon that uses 1.2GHz mobiles (with separate head units) and supports digital voice to other units. It also falls back to analog for the early adopters. More importantly, they also offer a base station that extends the range of the mobiles by linking to other base stations using a 10GHz backbone. Assuming the rackmountable base station supports integration into a colocated analog repeater, I would venture to say that this system is *precisely* what I recommended 3 years ago, and they even added a 128k data channel.
The big question is, of course, if some hams setup a "send a message to your family" station in Central Park when the blackout struck, which would have been better for getting messages out? HF needing only local power (and propogation), VHF needing local power and a repeater with power or reachable base station in South Jersey, or an AO-40 station with 1.2GHz and 2.4GHz yagis (and a 6-hour pass)?
500MHz is a lot of bandwidth to fill, and this new product will at least get SOMETHING up in there.
I have one of these these "mainframes on a card". I asked around the office if anyone had one, and a mainframe developer in Harrisburg wanted to get rid of one for a couple of years. So I drove out and picked it up. It runs OS/2 and has a real S/390 chip on a microchannel card. He had it way tweaked, and now I need to find out how to reload VM on it:( I got the installation media from someone at work, but he said it was for a newer version and would not work on the original P/390 I have.
It is a *totally* different kind of beast, but I really need to learn it.
Does anyone have a version 1 P/390 based on the PC Server 500, not the 330 with installation media? TiA
20,000 concurrent virtual machines running Linux on a single z990? Seriously?
Seriously. (Very) low utilization, though. Think "ISP in a box" outlined previously. How many people sign up for web space and dump a static page up there to show their friends and family? More realistically, consider wiping out an Intel and UNIX server farm. IFL, the Integrated Facility for Linux lets us charge much less to activate new engines on a z box that can *only* run Linux. Dedicate one engine to Linux images, and it caps out at around 25k.
how do the virtual machines on Z compare to the VMware VMs?
Not at all. The VMs on Z still see the mainframe hardware. Different CPU, disk and storage subsystems, network interfaces, everything is different. Beginners' skills still apply though, because the Linux VMs on Z still have disk, network, security, features, and CPU utilization that needs to be tuned, just like on Intel.
No you cannot migrate VMs. You have to get used to the fact that mainframes do not run or present Intel chips to the VMs. Windows (not counting CE) runs only on Intel these days. Linux can use just about any CPU architecture that was ever released, including the six used by corporate enterprises, Intel, AS/400, SPARC, Power, PA-RISC, and S/390.
once the hardware has been virtualized it doesn't matter what the back end
Be careful here. VMWare does in software what S/390 does in hardware. It presents each running copy of an OS a full copy of the CPU of that architecture. VMWare presents one (two as of August) Intel CPU to each VM, while VM on the mainframe uses the chip's own virtualization technology to present one or more S/390 to each running copy of an OS.
Just as you can run almost any Intel OS in VMs running under VMWare, you can run any mainframe OS in VMs under VM on the mainframe. I know I'm running on, but VM on the mainframe doesn't present an Intel emulator to each VM, it presents a copy of the mainframe CPU to each VM.
Slashdot Mirror
plus explaining to 130,000 kids how to rpm -Uvh rpmname.rpm isn't optimal
There are plenty of automated software management systems out there, commercial and free. At the very least, create a tiered package deployment infrastructure. Setup the laptops with a system based on RH9 and cron-apt package. Then central IT in Lansing tests and approves new RPMs to be pushed out to the students. They serve up the new RPMs on a centralized server, and each school district mirrors the site. Each laptop uses cron-apt to update itself from the local mirror during homeroom. You don't even have to consider running the Ximian updater, just apt for RPM and good ole cron.
Each laptop runs sshd and has a HelpMe! application which really just opens an IM window to a Jabber chat room hosted by Central IT. Worst comes to worst, someone from the central help desk ssh's into the kid's laptop to fix the problem, or the kid drags the dead laptop to the school's tech liasion. A central help desk of 5 or 10 and a lead sysadmin for 130,000 kids doesn't sound too expensive to me. Sounds pretty damn beautiful in fact. Remember, if an app (like a jabber client) does something you don't want it to do (like turn each laptop into a chat whore), you have the source and can actually do something about it.
If they have even a little money to spend they can use IBM Director or even CA's AMO and SDO (cough) for commerical inventory and software delivery products.
No sixth graders having to learn RPM, sorry. Nor any sixth graders getting their laptops infected by opening their email or visiting a malware site. I had a chance to drive the development of Linux laptops for 100,000 students nationwide. Too bad the organization ran out of money doing all the other stuff you need to do to run a school system.
HBO On Demand just started on Oct-1 and I can play any show or movie airing during the month at any time without incurring extra charges. Just like the commercial says, you can play, pause, fast forward and rewind any On Demand show right from the remote. Add that to my 144-hour TiVo, and its a crankin' setup. And those movies are separate from the virtual Blockbuster in my cable box with hundreds of titles for $3.99 and a 24-hour viewing window. So your nirvana may not be at hand, but having 500 movies and shows at my fingertips (and 1/3rd at no extra charge) is a pretty good start.
Each time this conversation comes up, I'm surprised no one mentions the Atom Films On Demand channel carried by Comcast. There's some cool stuff in there.
stream you the video of the concert to you[sic] PDA
I could have used something like at the performance of Chicago I saw recently. We sat so far over to one side, I had to duck down to peer below a stage light just to see the right half of the stage. But we had the privilege of paying the same price as the folks who sat front row center, so I was grateful for that.
this whole CLI vs. GUI thing is stupid, as long as we're using Linux who cares?!?!?!
n d. ..here
must...resist...asking...if...he's...new...arou
j/k
If anyone does it, the guys at LUFS will write a driver for exactly that. "Remember, everything is a file, and if it isn't, it should be!"
:)
I can see a driver that posts to encrypted anonymous groups optionally via anonymous remailers and then checks google for updates. Post 2 messages per chunk: a pointer to each chunk's subject line, and each chunk. The data in the pointer-posts should be enough to formulate a directory structure. (Hey, I might be onto something here...) Now, for the UI. Something that can generate posts and reconstruct files. Perhaps an ftp-like interface? Or just a CLI that can be backgrounded for the reconstruction phase.
It seems like that would give you much better access to vast data storage without having to staple together slow, medium, and fast-access data. If they do write drivers for all those methods, I vote for this to be the first one they tackle
I look at it this way.. you can "cron" up Windows Update or apt-get update; apt-get upgrade. However, Windows boxes will typically have to be rebooted and if you're running 24x7 you better have them clustered (with different auto-update times). And that goes wayy beyond all the stuff about trusting and automatically installing Windows patches.
At least on the Linux side, there's a little hope. Suppose you have 3 tiers.. you can setup a box that will keep debs or rpms updated via cron. apt-get update; apt-get upgrade -d IIRC. Then in the next tier, you have a sample of the different types of machines that can be found in your organization -- web, db, file, directory, whatever. Configure those boxes to point to a web site on your 1st level box in their apt sources. You can manually copy updated debs or rpms into the 'testing' website and let the beta boxes automatically update themselves using cron-apt. Test them all, check to make sure functionality is unaffected.
Then, set the last tier boxes (your production level boxes) to use a different virtual web server (say, "production") on the 1st tier box as their apt sources. Once you have tested and approved the patches applied to the 2nd tier boxes, you can copy the new debs or rpms to the DocumentRoot that serves the 3rd tier boxes. Et voila! Staged releases just by copying a few hundred k to a couple of directories on one box, spaced out over the course of a week. No more "patch money [sic]".
See cron-apt for reference. And yes, people, RH has apt in all its dependency-checking glory. So move along.
Debian really is the way...
stop it
stop it
stop it
stop it
STOP IT
PLEASE
sheesh
How soon before someone hacks it?
Back in January, I was staying at the Hilton Times Square, and while hurrying back inside one night, I looked up at all the monitors over top a neighboring shop, probably Madame Tussaud's Wax Museum. Sure enough, three of the monitors were displaying pop-up spam right over top of their regular advertising!
Who could imagine having that much tech and throwing it all away by directly attaching Windows boxes to the Internet without a firewall. Have they no shame!? Its been several months since I've looked, but I'm sure someone has had their way with those machines by now. Now THAT would be a fun box to 0wn. Pair that up with a laptop on wireless and you can have some serious fun with passers-by.
Sounds like I have a much smaller circle of friends. Remember, too, that when you put up your own Jabber server, you can still connect with your contacts on those privately owned systems. The difference is that if they prefer to use a "chat aggregation client" like Gaim or Trillian, you will be able to offer them an alternative when the owners of their preferred IM system throw the big red switch on 3rd party clients.
Jabber was not built to just switch everyone over to a new, privately owned network. It was designed from the beginning to allow groups, defined any way you want, to run their own server, with its own custom features, but still link to any outside networks with their own set of users, either Jabber, AIM, Yahoo!, MSN, or anything else.
Finally pushed me to install a Jabber server for my circle of friends. I've been using Gaim for a year or so, and its time to get my wife off Trillian ASAFP. I'm just sick of using closed source apps against a few servers who's owners occasionally show a propensity to piss all of us off.
Nice to have an alternative available that *I* control, end to end. And no, the Gaim Win32 interface does not suck. Just use the Wimp theme.
Exactly. I helped build NYU CompSci's very first web site and spent many days converting the technical paper collection to PS when electronically available and scanned to TIFF when it wasn't.. like for papers dating back to the late 60's.
There was some cool stuff buried in there.
SCO convinced Ren
Ren was always easy, it was Stimpy that was always a stickler for details.
Sorry.
I guess I *am* new around here, because I figured more people here would understand that the b2b uses of the 'net will make all the consumer-level stuff seem like a speck, a blip on the radar, miniscule and insignificant. A million dollars will be spent linking enterprises for every thousand dollars spent by consumers.
If you're only looking at web pages, you're missing 75% of the traffic that traverses the Internet today.
It doesn't need any of those privileges, but Linux has no mechanism to protect you on that level
/etc/shadow from everyone (even root) *except* /bin/login, who has read-only access.
Obligatory reference to LIDS. LIDS lets you specify exactly which users and applications have any combination of an extensive list of privileges, including read or write privileges at the file level and opening sockets. A common example would be hiding
My old quote used to be "can't get root if there is no root", but you weren't claiming that Linux suffers from multiple privilege escalation vulnerabilities. All told, you are right about native Linux, but there is at least one fix.
All told, I think "CW beacon" would have been the best choice. Many/most/all(?) satellites do transmit a CW identifier for a number of purposes, including the one the author used it for. He listened for the "I am alive" morse code transmission from the sat and heard it. So the sat didn't explode or unexpectedly fire its repositioning jets and move out of its assigned orbital slot.
CW = carrier wave
:)
Sorry, "continuous wave". Read the nick.
A 12 year old kid will NOT be able to compete in tomorrow's marketplace or the global community if she doesn't grow up Internet-connected today. If that mother disconnects the cablemodem, one more kid from the next generation will be lost.
But satellite tv has no redeeming educational value, and kids running around half naked eating dirt is just funny as hell.
I'm trying to get a quirky PC to boot Debian 3r1, and I can't get it to boot without the rescue floppy.
LILO gave L 40 40 (repeat indefinitely) at first. I added
device hda=0x80
to lilo.conf and that replaced the L 40's with just LI. Compiled a 2.4 kernel, and now I get
LILO
L
instead of
LILO
Uncompressing Linux......................
so I decided to try GRUB, and now the system boots to the grub CLI, and when I try to specify my 2.4 kernel, I get Error 13, unknown or unsupported kernel format. I'm currently recompiling the kernel as a.out format instead of ELF, but it'll take another hour.
Anyone have any tips? And no, the PC isn't bad, or the RAM, or the hdd's.. I've had this machine running for about four years.. it ran multiple versions of RH6 and 7 before they decided to require a minimum of 64MB RAM. Any help on getting GRUB working would be appreciated.
And if you stick around long enough at DeveloperWorks, you'll make a grid out of those PCs and offer the district more compute power than they ever realized they had.
I'd be more concerned with an attempt to take one out on takeoff than landing
That's what happened in Kenya several months ago. A SAM-7 was fired at a chartered Israeli jet that had taken off a few seconds earlier. The terrorist fired it off too early, and the SAM never picked up on the IR signature you worry about. The pilots felt a bump on the left side of the plane and saw the smoke trail as the rocket flew past the plane. Once they reached Israeli airspace, they were escorted in by Israeli F-15's and landed safely.
I'm surprised you don't seem to remember this event. Its exactly what you describe. Except, of course, the missile failed to destroy the target.. more like bounced off it.
I, like many here I would guess, have the BEFW11S4 (i think) wireless AP/router/4-port switch model. One thing that I would change right away is to make a "real" DMZ. I'm sure most people here know enough to not put a box in the "DMZ" which really just means "no firewall for this one box", but instead forward ports in to a specific box.
I would want to be able to firewall off an Internet-accessible box from the rest of my network while forwarding one or more inbound ports to that box. That would be better than all the port shenanigans you have to go through to support NetMeeting, for instance.
As a tech no-coder for ~10 years, I don't think HF is the critical band. Not to disrespect the very foundation of our hobby, but, IIRC, we have 10.0GHz to 10.5GHz and DirecTV has 10.5GHz to 11.0GHz. I believe we have the exact same bandwidth as them, and butt right up against them (IOW identical RF properties). I think its fair to say that they do a wee bit more with their allocation (which cost mega $$$) than we do with ours.
Back in y2k, the ARRL put up a website asking us for ideas as to where ham radio should go from here. They wanted totally new, uheard of ideas that may spark some interest by the manufacturers. I posted a warning that we need to focus on the 10GHz band so that we don't risk losing it. Remember HF is 30MHz wide, and we get, what 3MHz of it all told? We have spots in 1.2Ghz, 3GHz, 5GHz, 10GHz, 24GHz, and 46GHz and we have **500MHz** in the 10GHz band alone.
The technology I proposed was one that focused on VOIP using high speed data links in cells and linking cells together using even higher speed links. I suggested using a class B multicast subnet to "channelize" frequencies. Then, as clubs upgraded their cell-linking technologies, they could extend a "channel's" reach and offer additional channels to their local users. As an example, the most popular repeater frequencies in NYC could be extended all the way out to Rochester by using extenders in the 10GHz range, while clubs along the path could choose from the selection of channels they wish to carry. I specifically suggested keeping a separate head unit to allow individual users to upgrade the link technology and gradually utilize more channels at their location.
Sure enough, in the latest CQ VHF, I saw a product coming soon that uses 1.2GHz mobiles (with separate head units) and supports digital voice to other units. It also falls back to analog for the early adopters. More importantly, they also offer a base station that extends the range of the mobiles by linking to other base stations using a 10GHz backbone. Assuming the rackmountable base station supports integration into a colocated analog repeater, I would venture to say that this system is *precisely* what I recommended 3 years ago, and they even added a 128k data channel.
The big question is, of course, if some hams setup a "send a message to your family" station in Central Park when the blackout struck, which would have been better for getting messages out? HF needing only local power (and propogation), VHF needing local power and a repeater with power or reachable base station in South Jersey, or an AO-40 station with 1.2GHz and 2.4GHz yagis (and a 6-hour pass)?
500MHz is a lot of bandwidth to fill, and this new product will at least get SOMETHING up in there.
I have one of these these "mainframes on a card". I asked around the office if anyone had one, and a mainframe developer in Harrisburg wanted to get rid of one for a couple of years. So I drove out and picked it up. It runs OS/2 and has a real S/390 chip on a microchannel card. He had it way tweaked, and now I need to find out how to reload VM on it :( I got the installation media from someone at work, but he said it was for a newer version and would not work on the original P/390 I have.
It is a *totally* different kind of beast, but I really need to learn it.
Does anyone have a version 1 P/390 based on the PC Server 500, not the 330 with installation media? TiA
A single tornato
Anyone else read that as "a single tomato" and envision that famous Italian festival?
20,000 concurrent virtual machines running Linux on a single z990? Seriously?
Seriously. (Very) low utilization, though. Think "ISP in a box" outlined previously. How many people sign up for web space and dump a static page up there to show their friends and family? More realistically, consider wiping out an Intel and UNIX server farm. IFL, the Integrated Facility for Linux lets us charge much less to activate new engines on a z box that can *only* run Linux. Dedicate one engine to Linux images, and it caps out at around 25k.
how do the virtual machines on Z compare to the VMware VMs?
Not at all. The VMs on Z still see the mainframe hardware. Different CPU, disk and storage subsystems, network interfaces, everything is different. Beginners' skills still apply though, because the Linux VMs on Z still have disk, network, security, features, and CPU utilization that needs to be tuned, just like on Intel.
No you cannot migrate VMs. You have to get used to the fact that mainframes do not run or present Intel chips to the VMs. Windows (not counting CE) runs only on Intel these days. Linux can use just about any CPU architecture that was ever released, including the six used by corporate enterprises, Intel, AS/400, SPARC, Power, PA-RISC, and S/390.
once the hardware has been virtualized it doesn't matter what the back end
Be careful here. VMWare does in software what S/390 does in hardware. It presents each running copy of an OS a full copy of the CPU of that architecture. VMWare presents one (two as of August) Intel CPU to each VM, while VM on the mainframe uses the chip's own virtualization technology to present one or more S/390 to each running copy of an OS.
Just as you can run almost any Intel OS in VMs running under VMWare, you can run any mainframe OS in VMs under VM on the mainframe. I know I'm running on, but VM on the mainframe doesn't present an Intel emulator to each VM, it presents a copy of the mainframe CPU to each VM.