If you RTA, it sounds like based on current advances, in 10 years we'll be at the point where disks are so large 920 TB each) that access will have to become sequential (making them like tape today, access speeds not increasing as fast).
That would leave room for RAM to essentially become used for random access in the way the disks are used today and perhaps current cache on the CPU to be used more like RAM is today?
A lot of wire-speed net devices are starting to look like this, with their info stored in a non-volatile storage device, but loaded into RAM on startup and all "work" done in ram.
It's easy to image a whole chain reaction of purposes for devices slipping into other functions as a result of varying levels of technological advancement in them.
I've been using their web services for a while now for BooksUnderReview.com to grab review info and while the interface and XML over http are great, Amazon's web services servers have a tendancy to become overloaded 20-30% of the time. I finally had to write into my scripts to do extensive error checking to make sure we really did get a completely valid response before processing any of the info.
There are entire websites based around the premise of using dynamic AWS (Amazon Web Services links), but with the AWS unreliability most of us have found that "local pre-caching" works best. For places looking for product info in an easy to get fashion, AWS works great as long as you plan around it's non-available times.
Some of their restrictions are very annoying (1 second between accesses and content storage time restrictions), but understandable. Still, it's something I'd love to see more and more major sites using.
I havent used AIX, HPUX or OS/360/390/400. Does anyone know of their strengths and edge over other OSes?
Yeah, their main strength (from their maker's standpoint) is that you get to pay IBM $1000's per hour or HP hundreds per hour in consulting fees for the priviledge of using them.
Really, unless you must use software or hardware that only runs on/with HPUX or AIX, don't bother. Even IBM is getting away from AIX for a lot of stuff.
Some of the biggest issues we've had are integrating with other companies who are running the above. I spent two days at an IBM facility for a conference to design how we were going to communicate with a business partner user MQ Series, since that was their IBM influenced "standard". The end result was to use MQ for "small" messages and a DIFFERENT proprietary solution for "large" messages like image files. This all because they had an IBM AS/blah blah culture.
In contrast, we spent about 30 minutes each getting all the other partner companies setup with some SSH2/SCP2 scripts to communicate with our Solaris box (even if they were running NT/2000, just had to change the version of what we sent them) and they've proven to be much less of a headache.
In summary, simplicity and standardization in Solaris, Linux or *BSD, etc... enables a lot of time and cost savings that the AIX/HPUX/OS of the world have missed.
I had the same comment hit home to me and am also considering grabbing another piece of paper to hang on the wall and list on a resume.
However, I think I've hit on the "solution". It's Western Governors University, which is fully regionally accredited and does competency based classes.
In other words, if you already know it or want to learn it from a book or doing it, you can then just take a test and get the credit.
Reviewing the requirements, I figure I can handle one semester of tuition to learn the two or three things that I don't already know in their program and graduate.
Disclaimer: Please, this really isn't an attempt to start an OS flame war, so don't reply to this with your opinion about which OS is better!
The article leaves out the "other" main competitor for stability as seen on Netcraft's top uptime sites BSD.
Scalability and cost
on
Sun's Last Stand
·
· Score: 5, Interesting
Sun's current "low-end" tactic of trying to replace Linux with Sun on x86 is going to win a lot of converts. There are a lot of applications out there and companies that are used to Solaris and that installed base isn't going to just go away peacefully.
The biggest argument for converting servers to smaller x86 boxes has been scalability and cost. Linux is a popular way to do that, but many companies have been using various BSD variants as well because they are more comfortable with server vs. desktop oriented software. Sun will do very well in those areas with their new emphasis.
For a company that wants to keep their big hardware on Solaris for some stuff, it makes a lot of sense to standardize on Solaris for their cheap x86 servers as well.
Apparently ye olde slashdot filter removed the joke just after the . in the above post because it was written in the forbidden greater-than and less-than signs.
Meant to say that:
It's not any more irrelevent than IPV6 or (insert ISDN-style or.net-style standard or technology here).
Seriously, there will always be standards and technologies that make it from being in the right place at the right point of the implementation/budget curve and those that end up being skipped or never really fully implemented because it doesn't make sense for most to do so.
The end result of course is, if you didn't spend years on the standard yourself and your company isn't betting the farm on it, then: "Who cares?"
Another good reason to get DSL instead of a Cable modem if you can, not that the phone companies are much better about their part of DSL!
Perhaps someone can explain to them in no uncertain terms that people are tired of being assessed new taxes for government enforced monopolies instead of letting these services live and die on their own.
As long as you are already recording, you can manually stop it and go back. Not as nice as one-button smooth buffering, but with the storage much greater than a typical car trip listening to the radio a workable substitute.
For "work-time" recording to time-shift to "drive-time" it seems like it'd work for that pretty good as well.
Read it closer. As long as you are already recording, you can manually stop it and go back. Not as nice as one-button smooth buffering, but with the storage much greater than a typical car trip listening to the radio a workable substitute.
Every time I hear something on the radio that I want to replay, my hand makes the same motion I use for my PVR and I wonder why this product isn't out on the market already. Now it is... imagine, radio without commercials as long as you are "behind" the broadcast in time.
Microsoft "interoperability" with Unix generally consists of "re-using" code in their networking layer. Of course they'd be interested in a license from a company that claimed to own it. After all, if they need an excuse down the road why they share code with *nix, they can now point to their "license" from SCO and say it's legit.
The best part of the joke was slashdot linking to theregister.com...
It's all about the numbers...
on
LCD Price Fixing?
·
· Score: 5, Insightful
Many of the laptop makers either own their own monitor factories (Like Sony) or get incredible volume discounts doing their own importing (say, Dell).
Items that don't sell well in "retail" channels get a much higher mark-up to make up for the small volume. The same item in lots of 1000 or more over and over again will sell dirt cheap. Ever noticed the price per 1000 of your favorite cpu when it comes out?
It's a bit of a catch-22. When customers buy more via retail channels, the prices will come down. When the prices come down, customers will buy more...
Eventually the retailers will get there trying to compete with each other, but with "most" (me and you not among them) customers are perfectly happy with what's out there now, there isn't enough demand for a big retailer to start stocking larger quantities and begin the price death spiral we've grown to know and love about computer parts.
Are you basing your analyis of the planet warming up on temperature readings from satellite measurements of the atmosphere or from accumulated weather ground-station data?
If the former, would you provide a link or reference of some sort? If the latter, that's been discredited as being localized (not global) effects as more concrete and other heat-absorbing/radiating materials are built up around ground-stations near airports, etc...
I haven't seen any credible evidence of actual climate change. Since you sound somewhat reasonable in your post, perhaps you could provide some?
When those ads first came out we were in the middle of an upgrade to 9i RAC throughout our company. I'm pretty sure we single-handedly supplied Oracle with enough crash info to create their first few patch sets.
Our Oracle DBAs posted copies of the print "Unbreakable" ads near their desks so that we could all convulse with laughter every time we saw them.
Plenty of the Netapp shelves out there, with and without drives. Heck, I've got a couple of extra shelves with no drives right now. Anyone want to buy one?
I've been buying shelves filled with drives rather then just drives just because sometimes it's cheaper that way. You can get a shelf with 7 36 GB drives on the market for about $3-4K now. Of course, it helps if you also have a Netapp head... then you get their awesome file system, etc... as well. That's the best argument for not building your own to me... the Netapp filer's file system and utilities are light-years ahead of a "regular" OS's file system for mass storage.
I'd love to get my hands on some blade servers running Mandrake on AMD64-x86. Testing a web or app server config on this hardware+OS might make them the new winners of the most server bang-for-the-buck contest.
It's stuff like this that causes us sysadmin types who have had code "handed-off" to us to make sure we code review it ourselves for a risk assessment and then prevent anyone on the development team from having any access to the production or even QA environments.
When a developer decides that change-control is an "obstacle" and starts solving problems by connecting to production to try stuff, that just shows that they didn't do a good job of programming, error handling/logging and unit testing in the first place. If your sh*t breaks then you should have developed it so that it logs a relevent error message and fails gracefully. This will enable you to test and reproduce without having to screwup production every time you reproduce that bug that crashes the application...
When you are responsible for 100% uptime of a production system that is critical to the company's existance and its your ass awake in the middle of the night and on the line if it ever hiccups, let alone crashes, maybe you'll have more sympathy for procedures like "mandatory delays, sign-offs, handing over to sys admin staff, etc."
For every great developer we have in-house (and we have a few, one we even made an honorary sysadmin), we have three that don't even bother to unit test their code before checking it into version control and asking for it to be migrated. Part of this is time pressures because of idiotic management that sets deadlines first and THEN gathers requirements and creates an unrealistic project plan, but the rest is just because (I quote from our developer/honorary sysadmin) "Your code has no honor!"
Slashdot Mirror
If you RTA, it sounds like based on current advances, in 10 years we'll be at the point where disks are so large 920 TB each) that access will have to become sequential (making them like tape today, access speeds not increasing as fast).
That would leave room for RAM to essentially become used for random access in the way the disks are used today and perhaps current cache on the CPU to be used more like RAM is today?
A lot of wire-speed net devices are starting to look like this, with their info stored in a non-volatile storage device, but loaded into RAM on startup and all "work" done in ram.
It's easy to image a whole chain reaction of purposes for devices slipping into other functions as a result of varying levels of technological advancement in them.
I've been using their web services for a while now for BooksUnderReview.com to grab review info and while the interface and XML over http are great, Amazon's web services servers have a tendancy to become overloaded 20-30% of the time. I finally had to write into my scripts to do extensive error checking to make sure we really did get a completely valid response before processing any of the info.
There are entire websites based around the premise of using dynamic AWS (Amazon Web Services links), but with the AWS unreliability most of us have found that "local pre-caching" works best. For places looking for product info in an easy to get fashion, AWS works great as long as you plan around it's non-available times.
Some of their restrictions are very annoying (1 second between accesses and content storage time restrictions), but understandable. Still, it's something I'd love to see more and more major sites using.
I havent used AIX, HPUX or OS/360/390/400. Does anyone know of their strengths and edge over other OSes?
Yeah, their main strength (from their maker's standpoint) is that you get to pay IBM $1000's per hour or HP hundreds per hour in consulting fees for the priviledge of using them.
Really, unless you must use software or hardware that only runs on/with HPUX or AIX, don't bother. Even IBM is getting away from AIX for a lot of stuff.
Some of the biggest issues we've had are integrating with other companies who are running the above. I spent two days at an IBM facility for a conference to design how we were going to communicate with a business partner user MQ Series, since that was their IBM influenced "standard". The end result was to use MQ for "small" messages and a DIFFERENT proprietary solution for "large" messages like image files. This all because they had an IBM AS/blah blah culture.
In contrast, we spent about 30 minutes each getting all the other partner companies setup with some SSH2/SCP2 scripts to communicate with our Solaris box (even if they were running NT/2000, just had to change the version of what we sent them) and they've proven to be much less of a headache.
In summary, simplicity and standardization in Solaris, Linux or *BSD, etc... enables a lot of time and cost savings that the AIX/HPUX/OS of the world have missed.
I had the same comment hit home to me and am also considering grabbing another piece of paper to hang on the wall and list on a resume.
However, I think I've hit on the "solution". It's Western Governors University, which is fully regionally accredited and does competency based classes.
In other words, if you already know it or want to learn it from a book or doing it, you can then just take a test and get the credit.
Reviewing the requirements, I figure I can handle one semester of tuition to learn the two or three things that I don't already know in their program and graduate.
Disclaimer: Please, this really isn't an attempt to start an OS flame war, so don't reply to this with your opinion about which OS is better!
The article leaves out the "other" main competitor for stability as seen on Netcraft's top uptime sites BSD.
Sun's current "low-end" tactic of trying to replace Linux with Sun on x86 is going to win a lot of converts. There are a lot of applications out there and companies that are used to Solaris and that installed base isn't going to just go away peacefully.
The biggest argument for converting servers to smaller x86 boxes has been scalability and cost. Linux is a popular way to do that, but many companies have been using various BSD variants as well because they are more comfortable with server vs. desktop oriented software. Sun will do very well in those areas with their new emphasis.
For a company that wants to keep their big hardware on Solaris for some stuff, it makes a lot of sense to standardize on Solaris for their cheap x86 servers as well.
What's the point of stealing IPs to spam? Haven't these guys ever heard of wardriving for IPs?
These guys really need some serious technical help...
(Yes, not meant seriously for those law/spam enforcement types out there!)
Apparently ye olde slashdot filter removed the joke just after the . in the above post because it was written in the forbidden greater-than and less-than signs.
.net-style standard or technology here).
Meant to say that:
It's not any more irrelevent than IPV6 or (insert ISDN-style or
It's not any more irrelevent than IPv6 or .
Seriously, there will always be standards and technologies that make it from being in the right place at the right point of the implementation/budget curve and those that end up being skipped or never really fully implemented because it doesn't make sense for most to do so.
The end result of course is, if you didn't spend years on the standard yourself and your company isn't betting the farm on it, then: "Who cares?"
The solution is obvious. Just add an electric shaver attachment to your PDA phone and it will be exempt under 125.204 from having to be shut off....
Another good reason to get DSL instead of a Cable modem if you can, not that the phone companies are much better about their part of DSL!
Perhaps someone can explain to them in no uncertain terms that people are tired of being assessed new taxes for government enforced monopolies instead of letting these services live and die on their own.
Of course, it's "stimulus generalization". Don't you?
You can find more information at the Pavlov site!
As long as you are already recording, you can manually stop it and go back. Not as nice as one-button smooth buffering, but with the storage much greater than a typical car trip listening to the radio a workable substitute.
For "work-time" recording to time-shift to "drive-time" it seems like it'd work for that pretty good as well.
Read it closer. As long as you are already recording, you can manually stop it and go back. Not as nice as one-button smooth buffering, but with the storage much greater than a typical car trip listening to the radio a workable substitute.
Even better. I like the FM transmitter functions on the "Radio Personal Recorder". Too bad it looks like most of their models are currently sold out.
Every time I hear something on the radio that I want to replay, my hand makes the same motion I use for my PVR and I wonder why this product isn't out on the market already. Now it is... imagine, radio without commercials as long as you are "behind" the broadcast in time.
Time to start hunting the web for the best price!
Two words:
"Prior Art".
I had "Video on Demand" working on my C64 sometime circa 1983 in conjunction with a couple of VCRs.
Microsoft "interoperability" with Unix generally consists of "re-using" code in their networking layer. Of course they'd be interested in a license from a company that claimed to own it. After all, if they need an excuse down the road why they share code with *nix, they can now point to their "license" from SCO and say it's legit.
The best part of the joke was slashdot linking to theregister.com...
Many of the laptop makers either own their own monitor factories (Like Sony) or get incredible volume discounts doing their own importing (say, Dell).
Items that don't sell well in "retail" channels get a much higher mark-up to make up for the small volume. The same item in lots of 1000 or more over and over again will sell dirt cheap. Ever noticed the price per 1000 of your favorite cpu when it comes out?
It's a bit of a catch-22. When customers buy more via retail channels, the prices will come down. When the prices come down, customers will buy more...
Eventually the retailers will get there trying to compete with each other, but with "most" (me and you not among them) customers are perfectly happy with what's out there now, there isn't enough demand for a big retailer to start stocking larger quantities and begin the price death spiral we've grown to know and love about computer parts.
Are you basing your analyis of the planet warming up on temperature readings from satellite measurements of the atmosphere or from accumulated weather ground-station data? If the former, would you provide a link or reference of some sort? If the latter, that's been discredited as being localized (not global) effects as more concrete and other heat-absorbing/radiating materials are built up around ground-stations near airports, etc... I haven't seen any credible evidence of actual climate change. Since you sound somewhat reasonable in your post, perhaps you could provide some?
When those ads first came out we were in the middle of an upgrade to 9i RAC throughout our company. I'm pretty sure we single-handedly supplied Oracle with enough crash info to create their first few patch sets.
Our Oracle DBAs posted copies of the print "Unbreakable" ads near their desks so that we could all convulse with laughter every time we saw them.
Plenty of the Netapp shelves out there, with and without drives. Heck, I've got a couple of extra shelves with no drives right now. Anyone want to buy one?
I've been buying shelves filled with drives rather then just drives just because sometimes it's cheaper that way. You can get a shelf with 7 36 GB drives on the market for about $3-4K now. Of course, it helps if you also have a Netapp head... then you get their awesome file system, etc... as well. That's the best argument for not building your own to me... the Netapp filer's file system and utilities are light-years ahead of a "regular" OS's file system for mass storage.
I'd love to get my hands on some blade servers running Mandrake on AMD64-x86. Testing a web or app server config on this hardware+OS might make them the new winners of the most server bang-for-the-buck contest.
When a developer decides that change-control is an "obstacle" and starts solving problems by connecting to production to try stuff, that just shows that they didn't do a good job of programming, error handling/logging and unit testing in the first place. If your sh*t breaks then you should have developed it so that it logs a relevent error message and fails gracefully. This will enable you to test and reproduce without having to screwup production every time you reproduce that bug that crashes the application...
When you are responsible for 100% uptime of a production system that is critical to the company's existance and its your ass awake in the middle of the night and on the line if it ever hiccups, let alone crashes, maybe you'll have more sympathy for procedures like "mandatory delays, sign-offs, handing over to sys admin staff, etc."
For every great developer we have in-house (and we have a few, one we even made an honorary sysadmin), we have three that don't even bother to unit test their code before checking it into version control and asking for it to be migrated. Part of this is time pressures because of idiotic management that sets deadlines first and THEN gathers requirements and creates an unrealistic project plan, but the rest is just because (I quote from our developer/honorary sysadmin) "Your code has no honor!"