apparently you can't kick a recent red hat/centos without at least 512MB allocated to the vm, or you get nonsensical errors that don't indicate the problem.
also, the world of virtualization is very confusing. all these different types of vm's with different implementations and different limitations, and no real simple guide on how their actual use relates to one another, and when you need one thing or another. esxi isn't really free since you have to pay for a tool to configure it. their only linux support tool can't actually change any settings, you have to pay for esx to get that. (sure you can unlock the hidden ssh daemon and import an existing vm, but then you have to create all your vm's on your own machine? thats kinda retarded)
xen looks like the best bet for a free cluster of VM's, next to that openvz. i'm still trying to figure out how they work on a low level but all in all it seems you have to use a customized patched kernel and using the kernel you want to use or the patches you want may be quite difficult if not impossible.
Tuesday, May 12, 2009
Monday, May 4, 2009
thanks for all the coffee
Starbucks is like a number of retail outlets with an antiquated technological business model: charge people for internet access. It's not strange to many people who feel the convenience of getting on the web in the comfort of their favorite java dealer's boutique is ideal and worth a small hourly price. But for the rest of the world that just wants a brief look at their mail or news before heading out the door, belly full of bagels and iced latte, the price isn't worth the convenience. Thus many stores such as Dunkin Donuts and even Dennys have incorporated free wireless internet as a part of their business model to drive revenue by keeping customers planted in their seat and hoping they get hungry.
For the old model there is a price, and it is usually pretty standard in how it is addressed. You go to a store, you find wireless, you go to a website - and you are redirected to a portal demanding $10 an hour. Once receipt of your payment is confirmed, access controls in a proxy allow you to once again peruse the internet "fairly" unrestricted. This model works best in hotels or other establishments where it's preferable to pay a premium to get some kind of comfortable access to data services.
Being a hacker, I dislike the idea of paying to attain information. I don't think all information should be free per-se - some information is of limited scope and a sensitive nature, and i'd rather not everyone seeing all the "information" I access from my personal computer at home, for example. But in general I think paying for something the world has come to accept as "free access" just because it's a convenience some people find useful enough to be worth money to have access to isn't right. It's similar to how software is sold - it may cost the same amount to develop a piece of software that processes e-commerce payments as it does to develop an image-manipulating program, while one product is billed as far more expensive than the other, merely because people are willing to pay that much more simply to attain it. It's simply not fair to hold someone economically hostage because you have something they want or need, even though it really cost you nothing to develop it. This is how the rich become the super-rich in our excellent system of capitalism - exploiting people's weaknesses and benefiting from passing something off as more expensive than it really is.
So keeping this in mind, let me explain how you can get free internet access at Starbucks. There are several methods. The first two require an outside server to run a specific program, often which you must pay for unless someone has granted you a high level of access (and subsequently trusts you very much) or if someone else sets up the software for you.
The first piece of software is called iodine - this also requires name server entries to be set up for the domain (and yes you must own a domain) you will be connecting to. The theory is this: Most
- to be continued -
For the old model there is a price, and it is usually pretty standard in how it is addressed. You go to a store, you find wireless, you go to a website - and you are redirected to a portal demanding $10 an hour. Once receipt of your payment is confirmed, access controls in a proxy allow you to once again peruse the internet "fairly" unrestricted. This model works best in hotels or other establishments where it's preferable to pay a premium to get some kind of comfortable access to data services.
Being a hacker, I dislike the idea of paying to attain information. I don't think all information should be free per-se - some information is of limited scope and a sensitive nature, and i'd rather not everyone seeing all the "information" I access from my personal computer at home, for example. But in general I think paying for something the world has come to accept as "free access" just because it's a convenience some people find useful enough to be worth money to have access to isn't right. It's similar to how software is sold - it may cost the same amount to develop a piece of software that processes e-commerce payments as it does to develop an image-manipulating program, while one product is billed as far more expensive than the other, merely because people are willing to pay that much more simply to attain it. It's simply not fair to hold someone economically hostage because you have something they want or need, even though it really cost you nothing to develop it. This is how the rich become the super-rich in our excellent system of capitalism - exploiting people's weaknesses and benefiting from passing something off as more expensive than it really is.
So keeping this in mind, let me explain how you can get free internet access at Starbucks. There are several methods. The first two require an outside server to run a specific program, often which you must pay for unless someone has granted you a high level of access (and subsequently trusts you very much) or if someone else sets up the software for you.
The first piece of software is called iodine - this also requires name server entries to be set up for the domain (and yes you must own a domain) you will be connecting to. The theory is this: Most
- to be continued -
Friday, April 3, 2009
going retro: old firefox fixes new issues
Linux kinda sucks. There's many reasons (some you can read at The Linux Haters Blog and Why Linux Sucks), but for me it's always been firefox. Everyone claims it's just the best browser ever, but I think it's possibly worse than Internet Exploder.
First there's the memory issues. Since about version 2.0, firefox has been incredibly bloated and slow, introducing god knows what under the covers to soak up as much RAM as possible and eventually stop working. It was so bad for a while they had official notices of when the new versions would come out to fix this persistent issue. To this day you cannot browse around a few websites without 500 megs worth of shit clogging up your system.
Then there's the cpu. Try viewing 3 or more myspace band profiles in a recent firefox on a 3GHz Pentium 4 without it slowing down your whole system. (My load average soars up to 4 or 5 at least) Opera with the same pages loaded is much more manageable, and the same goes for most websites with flash audio and video. Is Opera doing some mystical incantations to make flash faster? I doubt it.
Today was the final straw. My system was left completely immobile by firefox (3). I opened a webpage and the VM went berserk, sucking up all available memory and swapping until the system was completely unresponsive. I rebooted instead of waiting around for the OOM to realize the system was fucked and kill the erroneous process.
To (hopefully) prevent this from happening again I have done the only thing I could think of to still use a halfway-decent browser while not blowing up my box: install an old version. I downloaded and packaged the 0.8 release of firefox (really still phoenix at that point) *without gtk2 or xft support* and the 1.0.8 release of firefox. I also have the latest version (3.0.8) on stand-by if I reeeally need a webpage which is Web 2.0-only, but so far only facebook has required it, and facebook sucks. /usr/bin/phoenix will point to 0.8 and /usr/bin/firefox-old/firefox-1.0.8 will point to 1.0.8, while /usr/bin/firefox goes to 3.0.8.
This should work for most cases as firefox 0.8 still uses the $HOME/.phoenix directory for user settings while 1.0.8 uses $HOME/.mozilla. I should really recompile 1.0.8 with a new user settings directory so I can use firefox 3.0.8 without worrying about clobbering old files in .mozilla, but i'm going to wait it out and see what happens. Alternatively I could run them in a chroot jail or just use the newest release of Seamonkey or the debian-inspired really-truly-free version of firefox.
The end result is fantastic. Firefox 0.8 page rendering times scream and the user experience is incredibly snappy and responsive (as a gtk1 app should be). Using firefox 1.0.8 is much the same as firefox 3, but it is noticeably faster at page rendering and much snappier in general use. "Modern" pages such as yahoo! mail work just fine, with only one or two minor page errors (which really could have been avoided with a better design choice).
My browser probably resembles swiss cheese from a security standpoint, but i'm willing to take the risk of someone exploiting my desktop to steal my incredibly sensitive camera photos that didn't make it onto my facebook. Lord knows the hacker underground has been after my flickr account's session cookies for years.
(psst... don't try looking for archived releases on Mozilla's main page. you'll have to dig through the ftp to find them: ftp://ftp.mozilla.org/pub/firefox/releases/)
First there's the memory issues. Since about version 2.0, firefox has been incredibly bloated and slow, introducing god knows what under the covers to soak up as much RAM as possible and eventually stop working. It was so bad for a while they had official notices of when the new versions would come out to fix this persistent issue. To this day you cannot browse around a few websites without 500 megs worth of shit clogging up your system.
Then there's the cpu. Try viewing 3 or more myspace band profiles in a recent firefox on a 3GHz Pentium 4 without it slowing down your whole system. (My load average soars up to 4 or 5 at least) Opera with the same pages loaded is much more manageable, and the same goes for most websites with flash audio and video. Is Opera doing some mystical incantations to make flash faster? I doubt it.
Today was the final straw. My system was left completely immobile by firefox (3). I opened a webpage and the VM went berserk, sucking up all available memory and swapping until the system was completely unresponsive. I rebooted instead of waiting around for the OOM to realize the system was fucked and kill the erroneous process.
To (hopefully) prevent this from happening again I have done the only thing I could think of to still use a halfway-decent browser while not blowing up my box: install an old version. I downloaded and packaged the 0.8 release of firefox (really still phoenix at that point) *without gtk2 or xft support* and the 1.0.8 release of firefox. I also have the latest version (3.0.8) on stand-by if I reeeally need a webpage which is Web 2.0-only, but so far only facebook has required it, and facebook sucks. /usr/bin/phoenix will point to 0.8 and /usr/bin/firefox-old/firefox-1.0.8 will point to 1.0.8, while /usr/bin/firefox goes to 3.0.8.
This should work for most cases as firefox 0.8 still uses the $HOME/.phoenix directory for user settings while 1.0.8 uses $HOME/.mozilla. I should really recompile 1.0.8 with a new user settings directory so I can use firefox 3.0.8 without worrying about clobbering old files in .mozilla, but i'm going to wait it out and see what happens. Alternatively I could run them in a chroot jail or just use the newest release of Seamonkey or the debian-inspired really-truly-free version of firefox.
The end result is fantastic. Firefox 0.8 page rendering times scream and the user experience is incredibly snappy and responsive (as a gtk1 app should be). Using firefox 1.0.8 is much the same as firefox 3, but it is noticeably faster at page rendering and much snappier in general use. "Modern" pages such as yahoo! mail work just fine, with only one or two minor page errors (which really could have been avoided with a better design choice).
My browser probably resembles swiss cheese from a security standpoint, but i'm willing to take the risk of someone exploiting my desktop to steal my incredibly sensitive camera photos that didn't make it onto my facebook. Lord knows the hacker underground has been after my flickr account's session cookies for years.
(psst... don't try looking for archived releases on Mozilla's main page. you'll have to dig through the ftp to find them: ftp://ftp.mozilla.org/pub/firefox/releases/)
Wednesday, April 1, 2009
some tricks to stay hidden
[1] If you're in a shell and typing some commands which you don't want to be preserved in your .bash_history, just run "kill -9 $$" when you're done with your shell. The parent process (shell) will be killed and not write anything to the .bash_history file.
[2] You want to log on to a box and use it interactively but don't feel like being monitored in the wtmp, use "ssh -v -T user@host" to log in. The verbose will tell you once you've successfully logged in. Now run 'last | head -n 5'. Your current login should not show up in the list. (Your login will probably still be logged by sshd in a separate file, but it's less noticeable)
[3] Encrypt your shell scripts: http://blogs.koolwal.net/2009/01/20/howto-encrypting-a-shell-script-on-a-linux-or-unix-based-system/
[3.1] Also use my script: http://themes.freshmeat.net/projects/encsh
[4] Encrypt all the files in a directory with cryptdir(1).
[2] You want to log on to a box and use it interactively but don't feel like being monitored in the wtmp, use "ssh -v -T user@host" to log in. The verbose will tell you once you've successfully logged in. Now run 'last | head -n 5'. Your current login should not show up in the list. (Your login will probably still be logged by sshd in a separate file, but it's less noticeable)
[3] Encrypt your shell scripts: http://blogs.koolwal.net/2009/01/20/howto-encrypting-a-shell-script-on-a-linux-or-unix-based-system/
[3.1] Also use my script: http://themes.freshmeat.net/projects/encsh
[4] Encrypt all the files in a directory with cryptdir(1).
Subscribe to:
Posts (Atom)