Discussion Locked

The Amazing Boring Blog

0 helpful
142 replies
Closed to new repliesThis discussion is closed. It stays readable, but it takes no new replies.

This is page 2 of the discussion. Read the opening post

142 replies

by jmgarvin

So, I don't know what to do next.  I was thinking of setting up a simple self CA SSL certificate in Linux, but I've no idea what I want to do.

If anyone has any suggestions, I'd be glad to hear them!

by rexworld

Not really a suggestion, just wanted to mention that I enjoyed the
squid pieces.  Nice little resource, gathering all that info and
links into one place.  The News.com folks are about to deploy
squid when they beta-test the new redesign, now at least I'll be able
to watch intelligently the config and deployment work they're
doing. 

Along those lines, there might be other proxy-type issues folks would
be interested in.  For example, here at CNET we use an
internally-developed proxy to hide the fact that most of our pages are
JSP's.  The proxy (among other things that it does) lets us hide
the .jsp extension with .html instead.  Useful so that we can
switch back-end technologies if we want without impacting people who
may have bookmarked our pages, that kind of thing.  If you know
how to do something like this with open-source proxy that might be an
interesting topic to present.

by apotheon

Ideas. Hmm. I don't know how much of this you might be able to help with, but I'll try to keep it down to areas about which you'll probably have some information:

  1. using rpm at the command line
  2. using YUM at the command line
  3. configuring YUM for the command line, including encryption configuration
  4. handling, managing, and installing software in tarball source
  5. the basics of console emacs use
  6. wireless network penetration testing
  7. madwifi installation and configuration

Do any of those sound like something you could take a whack at explaining? I know a little about most, but not a lot about any of them. Clearly, I'd like to know more.

by jmgarvin

Hmmmm...

First I think I'll take the first suggestion and finish up Squid with the obfustation factor(s).

Then, I'll tackle rpm and yum (basics) and then move on to advanced yum (it is a pretty powerful tool).


by pravat

Yaap, When SELinux is enabled it is very difficult to configure Samba,
take a example of Apache, this also will not work if you don't take the
Web Directory to the /var/www/***, becz it works with i the chrooted
mode. So, you have to find the directory which has been pre-specified
in the SELinux to share with samba else it won't work.

Bye

by jmgarvin

If you want to add yourself go there.  They actually have a
pretty good site setup and may have an answer to Palemto's question
about good IT blogs.

Check out my ranking (-134738275)

by macomolina

Hi,

"gracias", for your teaching, i run a RH8 with squid, but some friend make the config, now i undestand what he do in the config.

maybe you want to write about Remote Install Services, i am new to linux, and this topic is very interesting,

Marco

by jmgarvin

RPM is the RedHat Package Manager. 
This is similar to the "Add and Remove Programs" in Windows.  RPMs
work in a good number of Linux distributions.  Mostly in Red
Hat-alikes.  While RPMs can be manipulated using the GUI (Red Hat
-> System Settings -> Add/Remove Programs), it is usually far
easier and far faster to deal with RPMs at the CLI. 

So, what do I do?  First, let's take a look at a snippet from the man page.

RPM(8)                          
Red Hat
Linux                         
RPM(8)

NAME
       rpm - RPM Package Manager

SYNOPSIS
   QUERYING AND VERIFYING PACKAGES:
       rpm {-q|--query} [select-options] [query-options]

       rpm {-V|--verify} [select-options] [verify-options]

       rpm --import PUBKEY ...

       rpm {-K|--checksig} [--nosignature] [--nodigest]
           PACKAGE_FILE ...

   INSTALLING, UPGRADING, AND REMOVING PACKAGES:
       rpm {-i|--install} [install-options] PACKAGE_FILE ...

       rpm {-U|--upgrade} [install-options] PACKAGE_FILE ...

       rpm {-F|--freshen} [install-options] PACKAGE_FILE ...

       rpm {-e|--erase} [--allmatches] [--nodeps] [--noscripts]
           [--notriggers] [--repackage] [--test] PACKAGE_NAME ...

   MISCELLANEOUS:
       rpm {--initdb|--rebuilddb}

       rpm {--addsign|--resign} PACKAGE_FILE ...

       rpm {--querytags|--showrc}

       rpm {--setperms|--setugids} PACKAGE_NAME ...

   select-options
        [PACKAGE_NAME] [-a,--all] [-f,--file FILE]
        [-g,--group GROUP] {-p,--package PACKAGE_FILE]
        [--fileid MD5] [--hdrid SHA1] [--pkgid MD5] [--tid TID]
        [--querybynumber HDRNUM] [--triggeredby PACKAGE_NAME]
        [--whatprovides CAPABILITY] [--whatrequires CAPABILITY]

   query-options
        [--changelog] [-c,--configfiles] [-d,--docfiles] [--dump]
        [--filesbypkg] [-i,--info] [--last] [-l,--list]
        [--provides] [--qf,--queryformat QUERYFMT]
        [-R,--requires] [--scripts] [-s,--state]
        [--triggers,--triggerscripts]

   verify-options
        [--nodeps] [--nofiles] [--noscripts]
        [--nodigest] [--nosignature]
        [--nolinkto] [--nomd5] [--nosize] [--nouser]
install-options
        [--aid] [--allfiles] [--badreloc] [--excludepath OLDPATH]
        [--excludedocs] [--force] [-h,--hash]
        [--ignoresize] [--ignorearch] [--ignoreos]
        [--includedocs] [--justdb] [--nodeps]
        [--nodigest] [--nosignature] [--nosuggest]
        [--noorder] [--noscripts] [--notriggers]
        [--oldpackage] [--percent] [--prefix NEWPATH]
        [--relocate OLDPATH=NEWPATH]
        [--repackage] [--replacefiles] [--replacepkgs]
        [--test]

DESCRIPTION
       rpm is a powerful Package Manager, which can be used to build, install,
       query, verify, update, and erase individual software packages.  A pack-
       age  consists  of an archive of files and meta-data used to install and
       erase the archive files. The meta-data includes  helper  scripts,  file
       attributes,  and  descriptive  information about the package.  Packages
       come in two varieties: binary packages, used to encapsulate software to
       be  installed, 
and  source  packages,  containing  the source code
and

       recipe necessary to produce binary packages.

       One of the following basic modes must be selected: Query, Verify,  Sig-
       nature  Check, Install/Upgrade/Freshen, Uninstall, Initialize Database,
       Rebuild Database, Resign, Add Signature, Set Owners/Groups, Show Query-
       tags, and Show Configuration.

Let's start with a simple query to see what packages we have installed:
# rpm -qa

This queries for all installed rpm packages.  q for query and a for all

Let's say you didn't want a package, let's erase it:
# rpm -ev <package name>

This will erase the package and remove it from the system.  e will erase the package and v is verbose

Let's install a package:
# rpm -vih <package name>

This will install (i) a package with verbosity (v) and hash (h)marks.

If we need to update a package, this is no big deal either.
#rpm -Uvh <package name>

This will update (U) the package name with the newer rpm and give verbosity (v) and hash (h) marks.


Squid Part 3 is coming soon!!

by jmgarvin

Apotheon, the man page is pretty huge.  I can post the whole thing, but it is quite long.

Of course, the man page for yum is quite long too ;-)

by jmgarvin

Verbosity, in essence, tells you what is going on.  You don't have
to use this, but it does help let you know what is going on.

Since rpms at the command line are pretty quiet, the hash marks let you
see the install progress.   It is a progress bar, but with
###### in leui of a GUI progress bar.

Does that help?

by uberman

I've set up a samba share on my SLES 9 server. I wish to mount this share to 100 Win 2K servers on an AD Domain at boot up time and make it persistant. I did all the Kerberos key gen and trusts and mounting the shares works. But I can't get the shares to mount and persist on the servers when no one is logged on. Does anyone know how to do this?

by jmgarvin

I'm not entirely clear on what is happening.  Can you be a little more descriptive and see what we can figure out?

You might want to post this on Tech Q&A as well.

by jmgarvin

As if I don't have enough stress in my life.  My mom, sister, brother-in-law, and 6 year old nephew live in Houston.  My sister, brother-in-law, and nephew live in far north Houston (if anyone cares it is a place called Spring).  My mom lives right be NASA.

So, my mom has been evacuated and will be heading west a bit (to basically the far western edge of Houston...which should be safe), but is fighting traffic every step of the way.  My sister went to get more water from the store (they only had 2 gallons and with a small child she thought it would be a good idea to have about 4 gallons total in the house).  So she gets to the store and there are people BEATING EACH OTHER UP FOR FOOD AND WATER.  Yes, you read that right.  People were fighting for food and water.  People are freaking out and buying gallons and gallons of gas they won't need.  My sister said the lines to the gas stations lead on for miles. 

Why is it that something as simple as a little preparation turns into damn near a riot?  How many gallons of water do you REALLY need?  Probably 1 gallon per person to get through the worst of it.  What really bugs me is the total lack of humanity people are showing.  What has happened?  Where is the respect for your fellow man?

On that note: I'd like to suggest keeping an "Emergency Kit" in your garage or closet.

1) At least 1 gallon of water per person for 3 days.  Sure you need more, but water takes up a lot of room.  A gallon should last (with rationing) a person 3 days or more.  You should drink more water, but if it comes down to it, you have it on hand
2) Canned goods and a P-38 can opener.  The P-38 used to come with the old C-Rations.  Make sure you get veggies and low sodium items.  Also it might be a good idea to stock up on goods that don't need water or don't need to be heated
3) A first aid kit.  I keep one around because I'm clumsy, but many people don't have one.  A SAM Splint is a must!   This will help if someone breaks a bone and you have to set it yourself.  Sure it isn't the ideal situation, but broken bones happen and this can keep things stable.
4) At least 3 days of your medication.  Sure you have to rotate it out, but make sure you have it!
5) Make sure you have fire hoods.  Now, these need to be more accessable, but you need 'em!
6) Portable radio.  I like the crank ones, but you can get what you want
7) A tool kit.  You need to have screw drivers, a hammer, a mallet, at least a 12v drill, pliers, and wrenches/ratchet (to turn off your gas)
8) Garbage bags.  Not only can they hold your trash, but they can keep you dry in a pinch!
9) Moist toweletts.  Ok, I know...but I like to be clean.  I'd prefer a portable shower, but you take what you can get.
10) An emergency ladder if you have a multi-story home. 
11) A 5 gallon can of gas.  Not only can you top off your car in an emergency, but you can haul it with you to get those few extra miles.

I know it's cheesy, but it is something to think about!  Keep in mind common sense and a little planning can make any emergency a little easier to deal with.

by evolvemind

Your preparation advice is wise. My family has lived in several places
that were subject to various natural and human-generated crises. While
in the lower Puget Sound area of Washington State, we experienced two
very noticeable earthquakes. In the same location a wind storm blew a
mature Douglas fir tree onto our house, breakiing through the roof
tiles. In all cases we were lucky, but we were prepared with everything
on your list except the splint. That will be a good addition.

I would add, for those living abroad, have your passports, visas, birth
certificates, pet health certificates, other important documents, cash
or travelers' checks, and anything else you would need with you for
evacuation or other emergency international travel in a sturdy travel
portfolio that be grabbed in an instant. While living in Europe, we
kept ours in a fireproof box.

Mark H

by stress junkie

And now for something completely different:
I just saw your TQ&A answer for Jessie's boy names. Ha ha. No
points for you. I can't believe that she rejected anyone's suggestions,
but let's face it. Your suggestions for boy names were so amazingly bad
I can only thank the gods that you did not name me.

https://www.techrepublic.com.com/5208-11184-0.html?forumID=57&threadID=172506&start=0

Too bad I saw it four months late. I figured that I would suggest
giving her boy a girl's name: maybe Jaqui for instance. :-)

by jmgarvin

Hey, my suggestions weren't just bad, they were awful!  I tend to like Samhain though ;-)

'Course it can't be any worse than naming your kid "Evel"

by jdclyde

12)  Have a few flash lights in different locations through out
your home.  We have had a few times in the last five years where
power has been knocked out for one to five days depending on where you
live.  The State had cut back on it's forestry department, so when
he had strong winds come in, the trees came down and the state was in
bad shape.  I only lost power for the day, but I have friends and
family that were out for a WEEK.

13)  Batteries.  For your flashlights, as well as for hand
held games for the kids.  It helps them to deal with sitting in
the dark with nothing to do.

14)  Candles.  A few candles will light up your room better
and longer than a flashlight.  Just make sure they are in a proper
stand so they don't get knocked over and start a fire.

These are all things that I have needed and used in the last few years.  I ALWAYS have them on hand.

by jmgarvin

Ok, so it has been a long time since I posted. I've been really
busy. I've taken on extra tasks at work, plus I'm going to
school, and I have a bunch of family stuff happening all at once!
It seems like it is a never ending deluge of problems!!

So enough complaining, on to the show.

Yum is a command line updater for Red Hat and Red Hat-alikes.
Fedora and Red Hat use yum to update and install packages. Not
only is it quick, but it is very easy to use. Yum simply goes out
on the net and grabs packages from a server. You can define this
server or have it grab stuff right from Red Hat!

First, make sure you have the GPG keys:
# rpm --import http://fedora.redhat.com/about/security/4F2A6FD2.txt

The nice thing is that you don't need to mess with messy updaters or the klugy MS Update.

So how do I use yum? Well, let's just type yum at the command line and see what happens:
# yum

Usage: yum [options] < update | install | info | remove | list |
clean | provides | search | check-update | groupinstall |

groupupdate | grouplist | groupinfo | groupremove | generate-rss |
makecache | localinstall >

Options:
-c [config file] - specify the config file to use
-e [error level] - set the error logging level
-d [debug level] - set the debugging level
-y - answer yes to all questions
-R [time in minutes] - set the max amount of time to randomly run in
-C run from cache only - do not update the cache
--installroot=[path] - set the install root (default '/')
--version - output the version of yum
--rss-filename=[path/filename] - set the filename to generate rss to
--exclude=package to exclude
--disablerepo=repository id to disable (overrides config file)
--enablerepo=repository id to enable (overrides config file)

-h, --help - this screen

So the important words are list, install, update, and remove.
Since yum is a "package" manager of sorts, it is very similar to
RPM. All I need to know is what is installed on my system

So, let's see what is installed on my system:
# yum list > list.txt
# grep installed list.txt | less

You can see all the packages installed on your system! Pretty cool, eh? So, what about updates?
# grep updates list.txt | less

You can see all the packages that have updates available.

Ok, great...now what? Well, I can update a package:
# yum update libtool

This updates libtool and will get any dependencies you may
need. So, updating is easy! All I have to do is
yum update

Sure, but what about installing? You simply do the same thing!
# yum install emacs

This will install emacs on your system!

Great, but I hate emacs and you are wasting valuable space on my hard
drive, you say! I disagree that emacs is awful, but let's get rid
of it to make you happy:
# yum remove emacs

GONE!

But my package wasn't on the list!! No problem! You simply
need to add some lines to your yum.conf and import the proper gpg
keys. You can find the gpg keys here:
http://www.fedorafaq.org/#gpgsig

As usual, if you have any questions or issues, drop in a comment.

by jmgarvin

So, you might be thinking, how do I share out directories (folders) and files in Linux like I do in Windows?  The answer is simple, NFS.  NFS is a nice way of sharing out directories and files with little effort!  The bonus is that you can edit your fstab and automount the NFS shares!  Of course doing that is beyond the scope of this particular post.

NFS server (make sure you su - to root or this will be no end of headaches!)

1) Make sure nfs is installed
# cd /etc/init.d/
# ls nfs
- If NFS is not installed simply grab it via apt or yum or whatever
# yum install nfs
2) We must make a directory that we wish to share
# mkdir /shares
3) If we want it to start on boot we can change that in using
# ntsysv
4) Now edit the /etc/exports file using your favorite text editor
# emacs /etc/exports
/shares *(ro,sync)
- This means to allow /shares to be exported from the NFS daemon and it allows anyone to connect and it is read only.
5) Now, (re)start your nfs service
# service nfs start 
- Or the full path
# /etc/init.d/nfs start

NFS Client (you don't have to be root for this, but you have to be root if you wish to edit the fstab)

1) Create a directory to mount the shares directory on the remote machine
# mkdir /home/<username>/nfs/<server_name>
2) Now mount the share to the newly created mount point
# mount -t nfs <ip>:/shares /home/<username>/nfs/<server_name>
3) Now make sure it is mounted by listing what is in the directory
# ls /home/<username>/nfs/<server_name>
- If that matches with what is in the /shares directory on the server, you have mounted it!

As usual, if you have any questions, just post them and I'll see what I can do!

by jmgarvin

Ok, so I haven't posted much.  I'm trying to get a paper published
on my idea of securing that various applianced that you may be running
in your shop.

It is kernel level security and focuses on why the exokernel is the
ideal choice in many cases.  I'll post the presentation I've come
up with and, if possible, the full whitepaper.

by a.mcpherson

I have been working on an idea for an exokernel OS, KAOS, since december 2001.
I would like to hear what features you would like, basically I am doing a exokernel rewrite of NetBSD kernel, with a new gui/cli combo that is currently being prepared in java.
The security stuff comes from my custom made cryptography that is shaping up quite well.

by jmgarvin

It sounds interesting.  What have you done so far?  Drop me
an email and we'll discuss what we can do to collaborate and create
something cool ;-)

by jmgarvin

I'm currently teaching a MS security course, which all in all isn't a bad course.  It goes into good depth with Kerberos, IPSec, VPNs, and Active Directory "security."    The course actually has some good information about MS Kerberos and MIT Kerberos and it explains how Kerberos works in both the MS way and MIT way (not too much is different).  It also goes on to explain how to integrate IPSec across platforms.  It explains PAP and CHAP and all the ins and outs of PPTP and L2TP.  It goes on to basic VPN structure and how to implement it with Windows.  The nice thing is that it is pretty honest about the security that MS has and where they need to work.  However, in the process of teaching the course, I thought I'd show my students some of the MS "Get the Facts".  I fed it to them, without telling them what it was other than a case study.  I didn't want to lead them on, so I just let them read it and draw their own conclusions.

I grabbed the Nissan Motor's case study and media.  I also decided to show them the Ebay spot and that case study.  This turned into an interesting discussion.  My students, being very observant and quite bright, said "Where's the beef?"  They couldn't dig around enough to find any real information.  I point them to Microsoft's "Get the Facts" campaign, and was again asked "Why don't they have any deep data on here?" 

I thought about how to approach this.  I don't want to sound anti-Microsoft in the classroom and I typically try to explain to my students they always need to get the best tool for the job.  I also push my students to do research on products and explain to me why they would choose Product A over Product B.  However, my students were frustrated they it was very hard to quantify WHY MS over Linux or Linux over MS.  It was to describe why they should choose some email product over Exchange.  They couldn't find the hard data on the MS sites.  They had to dig around the net to get away from the marketing fluff to find real data and sometimes it was hard to find non-bias data.

My point is this:  Microsoft you need to put up hard numbers.  Just saying your product is better or putting up a case study that has little true content does not mean you have a better product.  My students can easily see past this and I would imagine that the IT consumer can as well.  Why don't you give us hard numbers?  Why don't you PROVE that I do indeed need to switch to your product.  Saying your product with save me $139 million dollars is meaningless unless you can explain WHY that is.  Saying the TCO of MS products is lower, doesn't make it so.  We need real data and some explanation as to why that is.  How do you quanitfy RIO?  How do we know?  There was very little information on how you came up with such numbers. 

At least my students learned one thing: The "Get the Facts" campaign is pure FUD.

by stress junkie

I don't know from your description whether your course goes on to ask
these students the kinds of questions that you presented at the end of
your post. It might be interesting to have them make the kinds of
decisions that CIOs, IT department managers, and system administrators
make all the time. Decisions like, in a new environment, why choose one
product over another. Then the same thing in an environment that is
heavily loaded with one product line already. Figuring out TCO,
problems with becoming a mixed environment, costs of remaining
dedicated to one product.

The kinds of problems involved could fill a couple of semesters. You
would probably never run out of new material. This is exactly the kind
of thing that I NEVER got in college. My whole curriculum was dedicated
to programming computers. It was really funny when, back in 1987, I
purchased Borland Pascal for MS-DOS and they had a single book that
covered by entire college curriculum. Algorithms, data structures, the
whole thing. In college I had no idea that the position of system
administrator existed. I've spent my career doing just that. So my
college experience had no relevance to my career.

But I'm not bitter. No sir. Not me. :D

by master3bs

I see what you mean. MS makes some pretty bold statements on that site; and some of them might actually be true.  I can see potentially how "although the initial hardware and software costs for a Linux server can be lower than those of Windows, higher expenses for application software and ongoing support raise its total cost of ownership to comparable levels."  But there is no further information.  If this is true MS needs to prove it.  If it isn't; they need another way to market it.

by apotheon

Gotta love the FUD. Most of it consists of picking out something
that varies widely, even if it tends very strongly against your favor,
and contriving some circumstance where it favors you instead. Then, you
hold that up as a case study. Voila. Instant marketing statistic.

Most of the MS anti-Linux FUD centers around administrative overhead and support costs. My experience, however, is as follows:

In a shop where 10-15% of the active systems run Microsoft Windows
(the rest Linux), 50-60% of my time and effort goes into
Windows-specific stuff, probably about 30-40% Linux-specific, and the
remainder platform-nonspecific.

The reason there isn't better statistical data in the "Get the
Facts" campaign is that it wouldn't favor Microsoft so well. In fact,
the results of many of the case studies in those ads have been
extensively massaged to get those Microsoft-supporting analyses. By
carefully tailoring case studies to bring about specific,
marketing-targeted results in limited focus areas, Microsoft gets what
it needs for its advertising, but there's a lot of additional data that
is generated as well -- much of it more damning to Microsoft's case,
and largely ignored in the ads. The actual study results can be dug up
in many cases, however, in pre-massaged state. I've done some of that
(exhaustingly difficult) research myself, and found that what's in the
ads is just the tip of the iceberg.

The biggest problem with the way these marketing/FUD campaigns are produced is in the fact that Microsoft is trying to maintain market share by proving competing solutions aren't good enough. If Microsoft would, instead, focus on proving its solutions are good, rather than that others are bad in comparison, I think it would have a better time of it in the long run. The long run doesn't look good on quarterly reports for the board of directors, though.

by jmgarvin

Stress,

Typically I do ask those questions (in a less biased way).  I also put my students in the position to explain to the CIO (me) why one OS over the other, given the information I would find. 

This benefits my guys two ways:

1) They have to think outside the box (and textbook) to find a solution to an unsolvable problem

2) They have to think like a sys admin and explain WHY a system functions a certain why and why it fits into the organizaiton.

That hardest part of the whole mess is getting them to recognize FUD on both sides (yes there is anti-MS FUD as well)

by aowen

HAYELP!!

OK, I've got Squid working on Debian Sarge, no probs. I've even setup an acl in squid to block access to one website..

So then I go looking for something with a blacklist, and come across both Squidguard and its webmin module.

First thing I do is download the blacklists...

Then my probs start...

I can see the contents of the blacklists via webmin, I can choose them in the ACLs but nothing seems to happen.

I've set up squidguard to block access to all sites and redirect to an
internal page from all but my pc. Which can go anywhere... I've set up
a destination group, and denied access to that, and nothing else, BUT I
can't get the blacklists to work....

Webmin has installed them into:
/var/lib/squidguard/db/blacklists

And I've just noticed a README in there... which has nothing to do with how to set it up...

Any Ideas?

Andy Owen

by jmgarvin

That is odd.  So the blacklist is allowing pass though of all the blacklisted URLs??  Can you post your config files?

by jmgarvin

First, let me start by saying that Sony is totally in the wrong and I
know this has been talked about ad nausium. Sony should have
NEVER produced a rootkit that took advantage of a system in such a
way. They not only betrayed the trust of the consumer, but
also took advantage of the unwitting and non-tech savvy. Shame on
Sony for doing this and shame on them for thinking that is was "ok" in
the name of anti-piracy.

On that note, MS should never have allowed its OS security to be
breached in such a way. This paves the way to all kinds of
intersting threats, the least of which is outlined at Security Focus.
If you really want to get sick, there are some intersting exploits
floating around the web that work a little too well. I won't post
links to them, but a savvy enough surfer can find millions of links in
seconds. We will also see a rash of rootkits that are very
similar to Sony's and they will have playloads that can execute quietly
because the rootkit is hidden (so the code is until it executes...and
even then if written properly it will still be quiet enough to be hard
to detect).

What really kills me is that Sony doesn't seem to think it is
such a big deal to create a rootkit to spy on users. If I was to
create a similar rootkit and somehow get millions to install it, I'd go
to jail. The worst thing to come out of this rootkit is
simplistic code hiding. Now I have a nice little area to hit all
my malicious code! Thanks Sony!

In the months and even years to come we will see exploits that take
advantage of this rootkit. My suggestion is this: Microsoft this
is your hour to shine. You should take this and prove you can do
security. Prove that your posture is that of a security minded
company. Stop this rootkit and cripple its ablity to be used in
further attacks. It is pretty simple MS, you know everything
about this. Come out like the white night and prove to the
consumer you CAN create a secure product.

So why the blog? I must say I am in awe of the audacity of Sony
to do this....I am also in awe of MS not having a clue...I just don't
know what to think.

by hwwgandolf

To think that MS. will stop the hole for this rootkit, is maybe a little optamistic. MS. and sony do what they want to make money. Whose to say that MS. didnt help them put it there in the first place. Money talks and bs walks. you know what i mean.

by jmgarvin

So, I had this vague idea for creating a unique key for port
knocking.  I'm currently hashing out the idea and I am going to
try to get a publication out of it.

However, three thing occured to me:
1) Is port knocking even really that useful when you can just tunnel though port 22 (SSH)
2) What are the long term benefits of having a secret knock to get in?
3) Even if there were a way to generate a key (as I say there is) that
creates a unique key that only the client and server know, have we
really gotten anywhere (even if the key is one time use)?

I honestly don't know.  While I think the idea is interesting, I
don't know if port knocking is useful.  Tunneling seems far more
efficient and locking down your firewall completely seems like a little
bit of over kill in some respects.

by stress junkie

So is this the end of new blog entries? Now that you are a published author I guess the old blog just doesn't have the prestige of writing an article.

by jmgarvin

lol, no I'm just so damn busy that I've not had time to blog ;-)

I'm going to try to start blogging again...I do miss it!

by jmgarvin

Most of you don't know this, but ATI and NVidia make native graphics
drivers for Linux.  In this blog entry, I'll discuss installing
the NVidia drivers.

It's a pretty simple matter to install them, although it can seem a
little difficult at first. If you follow my simple directions,
you should be just fine.

1) Read the EULA for the driver
http://www.nvidia.com/object/nv_swlicense.html

2) Download the drivers for your NVidia card (this is not for mobile devices)
32 bit architecture - http://download.nvidia.com/XFree86/Linux-x86/1.0-8756/NVIDIA-Linux-x86-1.0-8756-pkg1.run
64 bit architecture- http://download.nvidia.com/XFree86/Linux-x86_64/1.0-8756/NVIDIA-Linux-x86_64-1.0-8756-pkg2.run

3) Change to run level 3. What!!?? Ok, this isn't hard. To switch to run level 3 simply:
A) type init 3
B) reboot. At the grub boot load screen select your Linux OS and
press E (for edit). Drop down one entry (it should say something
like kernel vmlinuz-blablabla)...Press E again. At the end of
that line put the number 3.

Now simply type:
# sh NVIDIA-Linux-.run

Follow the directions. Let the installer update your X configuration.

Now your NVidia driver is installed! You can type:
# startx

And now your GUI works. If you don't see the NVidia splash screen, then you've done something wrong.

Simple!

by apotheon

Runlevel 3? According to whose version of runlevel 3? Why 3 in particular?

Are you saying it should be a CLI-only runlevel? That seems to be the logical option. Of course, I've seen Linux distros make runlevel 3 start up KDM or GDM.

Why do I think you're talking about Fedora Core's defaults again?

by jmgarvin

Ha, the quirks of Red Hat (and me not thinking) strike again!  I should probably quantify:

Run Level 3 in the Red Hat/Fedora world means multi-user, no GUI (or GUI log-in like GDM).

Run level 2-5 are all the same in the Debian world.  This means
multi-user and can imply GUI or not (depending on how the rc.d is setup.
http://www.debianhelp.co.uk/runlevels.htm is pretty clear on run levels.

So to clarify: Start without a GUI of any sort ;-)

Thanks for catching that Apo...Where would I be without you!  :-)

by stress junkie

Okay so this is months after the fact but let's do this anyway.

Apotheon, I know what you mean by "Whose run level 3?". I was very unhappy with Debian when I installed KDE and it made run level 3 run a GUI on startup. In honest-to-god-real-Unix run levels are not monkeyed with much. Run level 3 should be full system startup with text console. It makes no sense to have run level 3 and run level 5 act the same. When distributions redefine run level functions it only serves to confuse people and to make having different run levels useless. Linux was intended to be a work-alike operating system imitating Unix. The more a distribution gets away from that model the more confusing it is for everyone, especially newbies.

by marianne.popp9

I've just recently set up a Linux Desktop (Unbuntu distro), but so much of thier explanations for a newbie....is just way past me, I want to thank you for breaking the iptables setup of firewall.  I definitely understood this and it's going into my notebook of HELP!....LOL!!!

Thanks again

This discussion is closed to new replies.