For those folk who continue to RISC vs CISC debate you'll find that both design mentalities have been munged together to where modern CPU's show qualities from both camps. Best description comes from this peice on arstechnica.... It's well worth the read.
Hmm, 8TBytes a days isn't going to be cheap. Gonna need lots of tape storage, lots of it.
I would recommend an IBM 3494 tape library loaded with about 6 3590 E1A drives (2 for write, 2 for read, 2 for reclamation). The 3494 is very flexible and modular, fully decked out it has a 6,000+ tape cart capacity. This would equate to approx 748TB of total storage capacity.
Now for managing the entire thing, I would use Adstar Storage Manager. ASDM is now sold by Tivoli and not IBM direct, but is the same product. Use the HSM component. Basically you'd have a huge filesystem. When the video data is streamed in it gets written to disk first (requiring about 8TB of disk). The HSM manager will then migrate the data off disk and out to tape, but still leave the file entry in the filesystem.
From that point on if you need to access the video, you open the file that appears to be in the filesystem, but the HSM runs off in the background and mounts the tape and starts streaming the data back to disk for your client to read. Your client will block waiting for the tape to mount, but that would be acceptable considering the amount of data being stored.
This post is incomplete, but you may want to follow up on the 3494's at
http://www.storage.ibm.com/hardsoft/tape/3494/in de x.html
Also, investigate ADSM and how its HSM works for using a tape library to back a disk subsystem.
http://www.tivoli.com/support/public/Prodman/pub li c_manuals/td/TSM/SH26-4083-01/en_US/HTML/cp5ech15. htm
http://www.tivoli.com/products/index/storage_mgr /
Out of curiosity I'd like to know what your budget is on this:)
Software Upgrade Paths
on
IBM Wants Linux
·
· Score: 2, Insightful
The idea of IBM dropping AIX 100% in favor of Linux is a pretty long shot. As long as they have paying customers for AIX support, AIX will continue to live. Now where Linux comes in as a big play for IBM may have something to do with upgrade paths. Say for instance company X developes an application but they can't afford to ramp to big iron hardware to run it. IBM sells them some netfinities running Linux to get them jumpstarted. Then if their business starts to expand they would have the ability to migrate them up to a RS/6000 or AS/400 based system. The big kicker is that they can maintain 100% portability across the hardware platforms. Migration is a simple compile away:) This is a pretty powerful proposition, especially with the market in its current state. VC is dry, revenues are down, the idea of starting cheap and ramping up when needed may be Linux's biggest strength.
~
Here where I work (large university), we're striving to prototype to just that. It's called eduPerson and it's a standardized LDAP schema that represents a unique person within an educational institution. When deployed, it will be cross linked to other LDAP servers similar to how DNS is organized. This will allow for a global directory across all participating universities/educational institutions. Public keys are also issued to all participants providing a global PKI (public key infrastructure) to be also be developed. As with other standards that started their roots in the university/educational setting, perhaps this one could spread to the private sector as well.
I would speculate that the 2.5 kernel would provide support to both statically compiled security models (those accepted into the mainstream kernel) and modular
compiled security models. Much like the current system allows you to compile a device driver statically or as a module. One of the major benefits from allowing the security models to be compiled/loaded as modules is for development reasons. It can become a bit problematic to have to recompile a complete kernel then reboot (or boot a vmware session) after changing the code. You might end up spending a lot of time staring at the screen vs. writing/testing code. By allowing modules, you can update the code, recompile the module, load the module, test, unload the module, repeat. In the end, it can save a lot of time. This has been a major advantage for kernel developers in the past and I would imagine it will continue for the developers working on the new security features.
Hmmm, so the magical debate over CISC vs. RISC rolls on. As I much agree that this machine would be a nice machine to have, I don't think that the $1k price tag is really worth it. Reason being that high end Athalon and Pentium systems can be obtained at a much more attractive price/performance ratio.
Hmmm, now how about raw performance? Granted the Sparc II is a nice CPU, but can it actually compete with an Athalon at 2x the MHz. If so, can it compete in such a wide margin that would justify the price. Check the Spec bench's...
Granted, these shouldn't be taken as the ultimate in performance, but I don't see a staggering lead.
As for those in the RISC vs. CISC camps. I hate to inform you that the RISC, CISC is all but dead. Current RISC designs now longer embody the RISC philosophies of days past. CISC cores blend in to the point that one couldn't distinguish it from its called RISC counterparts. Modern CPU's are cutting the edge of new design concepts. If you feel the need to follow up on this. I suggest reading the following...
Actually Linux's stack has gone through a complete overhaul in 2.4 that would make it compete quite nicely against that of BSD's. It now has nifty features such as wake on one. It's also now a multi-threaded stack which has been written in an unserialized fashion to allow for good scalability. By far it'll run leaps and bounds better than the stack in 2.2. The big boys better start to take notice.....
Had this same exact problem a while ago after compiling an installing glibc 2.2. My solution was to check out the latest CVS release of gcc 2.95.x, which already had the fix. Even recompiled glibc 2.2 with the gcc CVS release with no problems. Also, haven't had a single incompatibility with the previous glibc 2.1. Perhaps, those folk, can give some specific occurrences of problems they've encountered.
I used to work for a large financial institution where they were attempting to deploy a rather larger customer service workbench application in Java. It was to replace a current system developed in VB. After two years of development, the project was scraped because they finally realized that they couldn't get an acceptable level of performance out of the apps. This is were the possibility of a 50% performance hit will kill you. They couldn't justify upgrading 400+ PC's to PIII 500 256MB specs to run the apps effectively.
Performance oriented software (AKA OS's, etc.) need to be close to the hardware to extract every ounce of performance. This is where C is most prevelant. On a side note, does anyone know what language all the commercial JVM's are written in?
Slashdot Mirror
For those folk who continue to RISC vs CISC debate you'll find that both design mentalities have been munged together to where modern CPU's show qualities from both camps. Best description comes from this peice on arstechnica.... It's well worth the read.
Article
Hmm, 8TBytes a days isn't going to be cheap. Gonna need lots of tape storage, lots of it.
n de x.html
b li c_manuals/td/TSM/SH26-4083-01/en_US/HTML/cp5ech15. htm
r /
:)
I would recommend an IBM 3494 tape library loaded with about 6 3590 E1A drives (2 for write, 2 for read, 2 for reclamation). The 3494 is very flexible and modular, fully decked out it has a 6,000+ tape cart capacity. This would equate to approx 748TB of total storage capacity.
Now for managing the entire thing, I would use Adstar Storage Manager. ASDM is now sold by Tivoli and not IBM direct, but is the same product. Use the HSM component. Basically you'd have a huge filesystem. When the video data is streamed in it gets written to disk first (requiring about 8TB of disk). The HSM manager will then migrate the data off disk and out to tape, but still leave the file entry in the filesystem.
From that point on if you need to access the video, you open the file that appears to be in the filesystem, but the HSM runs off in the background and mounts the tape and starts streaming the data back to disk for your client to read. Your client will block waiting for the tape to mount, but that would be acceptable considering the amount of data being stored.
This post is incomplete, but you may want to follow up on the 3494's at
http://www.storage.ibm.com/hardsoft/tape/3494/i
Also, investigate ADSM and how its HSM works for using a tape library to back a disk subsystem.
http://www.tivoli.com/support/public/Prodman/pu
http://www.tivoli.com/products/index/storage_mg
Out of curiosity I'd like to know what your budget is on this
The idea of IBM dropping AIX 100% in favor of Linux is a pretty long shot. As long as they have paying customers for AIX support, AIX will continue to live. Now where Linux comes in as a big play for IBM may have something to do with upgrade paths. Say for instance company X developes an application but they can't afford to ramp to big iron hardware to run it. IBM sells them some netfinities running Linux to get them jumpstarted. Then if their business starts to expand they would have the ability to migrate them up to a RS/6000 or AS/400 based system. The big kicker is that they can maintain 100% portability across the hardware platforms. Migration is a simple compile away :) This is a pretty powerful proposition, especially with the market in its current state. VC is dry, revenues are down, the idea of starting cheap and ramping up when needed may be Linux's biggest strength.
~
Dumb Mega-Corporation Accomplishment?
Here where I work (large university), we're striving to prototype to just that. It's called eduPerson and it's a standardized LDAP schema that represents a unique person within an educational institution. When deployed, it will be cross linked to other LDAP servers similar to how DNS is organized. This will allow for a global directory across all participating universities/educational institutions. Public keys are also issued to all participants providing a global PKI (public key infrastructure) to be also be developed. As with other standards that started their roots in the university/educational setting, perhaps this one could spread to the private sector as well.
http://www.educause.edu/eduperson/
Uh uh, and did you bother to look at other profiles. Say, perhaps, Oracle, Sun, JD Uniphase, Juniper...... blah
I would speculate that the 2.5 kernel would provide support to both statically compiled security models (those accepted into the mainstream kernel) and modular compiled security models. Much like the current system allows you to compile a device driver statically or as a module. One of the major benefits from allowing the security models to be compiled/loaded as modules is for development reasons. It can become a bit problematic to have to recompile a complete kernel then reboot (or boot a vmware session) after changing the code. You might end up spending a lot of time staring at the screen vs. writing/testing code. By allowing modules, you can update the code, recompile the module, load the module, test, unload the module, repeat. In the end, it can save a lot of time. This has been a major advantage for kernel developers in the past and I would imagine it will continue for the developers working on the new security features.
Errrr, momentary lapse of reason...Too much coffee today :)
s/athalon/athlon/g
Hmmm, so the magical debate over CISC vs. RISC rolls on. As I much agree that this machine would be a nice machine to have, I don't think that the $1k price tag is really worth it. Reason being that high end Athalon and Pentium systems can be obtained at a much more attractive price/performance ratio.
Hmmm, now how about raw performance? Granted the Sparc II is a nice CPU, but can it actually compete with an Athalon at 2x the MHz. If so, can it compete in such a wide margin that would justify the price. Check the Spec bench's...
2000 Integer Results
2000 Floating Point Results
Granted, these shouldn't be taken as the ultimate in performance, but I don't see a staggering lead.
As for those in the RISC vs. CISC camps. I hate to inform you that the RISC, CISC is all but dead. Current RISC designs now longer embody the RISC philosophies of days past. CISC cores blend in to the point that one couldn't distinguish it from its called RISC counterparts. Modern CPU's are cutting the edge of new design concepts. If you feel the need to follow up on this. I suggest reading the following...
RISC vs. CISC: the Post-RISC Era
and to track the history of your favorite CPU...
Here is a good place to start..
Hmmm, if you have a few bucks to spare, pick one up. But I don't see it as a vastly superior platform.
Actually Linux's stack has gone through a complete overhaul in 2.4 that would make it compete quite nicely against that of BSD's. It now has nifty features such as wake on one. It's also now a multi-threaded stack which has been written in an unserialized fashion to allow for good scalability. By far it'll run leaps and bounds better than the stack in 2.2. The big boys better start to take notice.....
Dude, do me a favor.....Shut the Fuck up.
Thank You
Can't get a response from Tom's site. Perhaps he's gonna have to rush that server into service a bit sooner than he thought :)
Had this same exact problem a while ago after compiling an installing glibc 2.2. My solution was to check out the latest CVS release of gcc 2.95.x, which already had the fix. Even recompiled glibc 2.2 with the gcc CVS release with no problems. Also, haven't had a single incompatibility with the previous glibc 2.1. Perhaps, those folk, can give some specific occurrences of problems they've encountered.
I used to work for a large financial institution where they were attempting to deploy a rather larger customer service workbench application in Java. It was to replace a current system developed in VB. After two years of development, the project was scraped because they finally realized that they couldn't get an acceptable level of performance out of the apps. This is were the possibility of a 50% performance hit will kill you. They couldn't justify upgrading 400+ PC's to PIII 500 256MB specs to run the apps effectively. Performance oriented software (AKA OS's, etc.) need to be close to the hardware to extract every ounce of performance. This is where C is most prevelant. On a side note, does anyone know what language all the commercial JVM's are written in?