Wednesday, 12 March 2014

Server 2008 - Setting Up a Domain Controller Part 1


Normally I stay away from the freedom hating Microsoft operating system but I thought I'd better er log this as its a bit of a pain and if you get it wrong it can be a B*****d to fix later in the life cycle of the network. The first thing to do is install Windows Server, Im using Server 2008R2 here, once installed, and you have set up the new password etc etc and your up to the desktop, get a Windows client and plug it into the machine for testing. Im using virtual box to do this as it would be a pain to set up a load of servers, however feel free to use real hardware...

1) Go to start and computer and give the computer an easy name, windows will call it something like WIN-xxxx12341564 so change it to, example-server or something easier. I'm calling mines mr-server as I don’t believe there is anything female about a windows server. Once you change the name restart the machine.

2) Set IP addressing - on the domain controller set this to a static IP address in this case, 192.168.1.254 with a 255.255.255.0 mask. The go to start and type dcpromo, and then run DC promo.exe. This IP address will become the IP of the domain controller for the client machines.

3) Once dcpromo starts, choose to create a new domain in a new forest, give your controller a name like mydomain.example, choose to install DNS and set a password for the domain and then finish the wizard, then restart the machine.

4) In server manager add roles and add file services and DHCP, set the IP address of the prefered DNS server to your server address (192.168.1.254) add in a scope for you server, in my case (192.168.1.1 - 192.168.1.100)


5) Now you can log into the domain, test the settings, go to your windows client machine: start|computer|system properties|change name. Give the computer a name, im calling mines mr7-client and set the domain to the name of your domain controller, (mydomain.example). This should then prompt you for a username and password. Enter Administrator and then your domain password. If you get errors here, go to the network and sharing centre and set the primary DNS server address to your servers address and then reboot the client. All going well you should get a message welcoming you to the domain.

6) Now its time to automate the adding of users to the active directory, using powershell, on the server start powershell and type:

Import-Module ServerManager

Add-WindowsFeature PowerShell-ISE

this should succeed. Now when you go to start you can open powershell ISE.

If you Type 

Get-ExecutionPolicy

into powershell it will say restricted, this means you cant execute scripts until you change the settings, this can be done by

Set-ExecutionPolicy Unrestricted

Unless you wnat to start self signing certificates...

Powershell is ready, start the ISE and lets begin writing out script, firstly we need to creat a CSV file with some users to add to the Active Directory.


we can get into the powershell in part 2...





Convert a machine from Kubuntu to Debian and Keep Your Home Folder

This will take a while, follow each step carefully and adapt it to your system

Part one, split the disks and install the OS

How to downgrade or move sideways from kubuntu to debian and keep your home folder, you will need a usb stick

  1. get unetbootin and gparted installed and then update your packages
    2)get an empty usb stick
3) open up Gparted, now format the usb stick to fat32
  1. get the small debian image from the debian website for your architecture (probably x86 or ADM 64)
5) Using Gparted format the usb stick to ext4
  1. pack everything you want to keep into your home folde
  2. Download a new gparted iso
  1. start UNETBootin and creat a bootable usb drive from the gparted iso
  1. now boot into gparted from the BIOS on the main machine and shrink your HDD down to provide space for the new operating system – in this case I have a 160GB HD so I am going to shrink the kubuntu partition to 100Gb and give the Debian install 60GB – be patient as this can take a while, do not try to abort this once you have started it.
  2. Once the HDD partition has been created, boot back into kubuntu and using UNETBOOTIN, now create a bootable debain USB stick get the small debian image from the debian site if you have a decent internet connection, to save time, you can get the updates during the install.
  3. Shutdown the pc and boot to the debain installer, its all easy stuff here keyboard layout, username etc etc- the important bit is when you get to the partition disks section, make sure you set the mount point of the new operating system ( / ) to the correct partition, and after the packages have installed choose yes to install GRUB to the MBR
  4. after the install remove the USB stick and you shoulkd be presented with GRUB, showing both kubuntu and debian, boot into debian and choose, you should now be presented with the new deaktop environment – the work is almost done, the next stage is setting up debian and migrating your home folder to the new operatings system...


Part 2 Set up debain

Okay – the first time you boot up Debian, if you are lucky GNOME3 will load straight away – no cigar pour moi this time, we are actually going to have to get the desktop environment working before we do anything else. I have an Nvidia Geforce GPU, 2GB however the graphics driver isnt in the default repo as its non-free (negative in the freedom dimension)

  1. You need to get the non-free sources from the backports, and the contrib ones – you willl have a system that is entirely free software at the moment which, however positive in the FDMN (vRMS) is going to be inferior in graphics (superior in ethics nonetheles)
  2. Go to applications | accessories | terminal and type su, stick in your root password, now type nano /ete/apt/sources.list
    add in>>>> deb http://ftp.uk.debian.org/debian wheezy-backports main
    and add the words contrib non-free to the end of your main ftp mirror so it looks like
    deb http://ftp.uk.debian.org/debian/ wheezy main contrib non-free
  3. run an apt-get update and apt-get upgrade and you should now have access to the non free sources which will allow the ionstallation of the nvidia drivers
  4. to get the drivers enter, at this stage go to terminal preferences and change to colour scheme to green on black (easier to see)
    apt-get install module-assistant nvidia-kernel-common
    m-a auto-install nvidia-kernel${VERSION}-source
this should get the matrix clicking away on some stuff, now get the drivers

    apt-get install nvidia-glx${VERSION}

you should get something alsong these lines

Conflicting nouveau kernel module loaded │
│ │
│ The free nouveau kernel module is currently loaded and conflicts with │
│ the non-free nvidia kernel module. │
│ │
│ The easiest way to fix this is to reboot the machine once the │
│ installation has finished.

Okay to reboot now but before you do take a picture of this with your phone as you will have to work without an x server to do the next stage so now windows or graphics

add this section to /etc/X11/xorg.conf

Section "Module"
Load "glx"
EndSection

Section "Device"
Identifier "Video Card"
Driver "nvidia"
EndSection

write out and then reboot the machone, when it restarts you should have gnome3 (-:

now in the terminal a grep of the driver should give you the correct one, instead of the free noveau driver

dpkg -l|grep nvidia-glx|awk {'print $3'}

Now we have graphics, next stage is to get flash working so you can watch videos on youtube, play music etc etc.

1) apt-get install flashplugin-nonfree
  1. update-flashplugin-nonfree –install
  2. apt-get update && apt-get upgrade

restart the browser and your up and running


Now we need to get sound working, this can often be a slipery one
install these one line at a time

apt-get install pulseaudio libao4 paprefs libpulse-mainloop-glib0 \
pulseaudio-module-jack pavucontrol pulseaudio-module-x11 \
gstreamer0.10-pulseaudio pulseaudio-utils libasound2-plugins \
paman pulseaudio-module-gconf libgconfmm-2.6-1c2 libpulse-browse0 \
pavumeter libglademm-2.4-1c2a pulseaudio-esound-compat libpulse0 \
libpulse-dev pulseaudio pulseaudio-esound-compat \
pulseaudio-module-gconf  pulseaudio-module-x11  \
pulseaudio-utils lib32asound2 lib32asound2-plugins  ia32-libs-gtk 
apt-get update and upgrade

then

usermod  -a -G pulse,pulse-access username
where username is your name
now make this config file 

/etc/asound.conf
pcm.pulse {
    type pulse
}
 
ctl.pulse {
    type pulse
}
 
pcm.!default {
    type pulse
}
 
ctl.!default {
    type pulse
}

reboot, then open a long track and get pulseaudio control open, set the config until you see the sound waves moving up and down, you may also have to tweak stuff in the alsamixer to get it working
Okay to get debain to see your phone when you plug it in (mines a samsung galaxy), also get folder sharing configured and stuff, you may notice that debian puts a (Public) folder into the /home/username/* directory by default – in this instance we can use this to share documents with other clients on the LAN.

Get these packages

apt-get install libcups2 samba samba-common
remove the "#" at the beginning of the line security = user
in the etc/samba/smb.conf file

restart samba


root@HOSTNAME:/home/michael# chown -R root:users /home/michael/Public
root@HOSTNAME:/home/michael# chmod -R ug+rwx,o+rx-w /home/michael/Public
root@HOSTNAME:/home/michael#


/etc/init.d/samba restart 

now back to the samba config file and point to the public folder 
[...] 
[allusers] 
  comment = All Users 
  path = /home/michael/Public 
  valid users = @users 
  force group = users 
  create mask = 0660 
  directory mask = 0771 
  writable = yes 


restart samba again

now add yourself to the users group

useradd michael -m -G users
smbpasswd -a tom
Now you can browse to the public folder from the other clients on the LAN

Files are now shared





Monday, 18 November 2013

Data Communication Systems


a very boring collection of notes that relate to data communications, any errors, email me, 'how dark the days were when memorising this information was the object':

 
LO1 Data Communications 'the process of accurate transmission of data from source to a receiver'

Analogue Can any value, accuracy is limited by the max & minimum values of the measuring instrument.
Digital Discrete number of fixed values, binary is the most common system.
Electromagnetic A periodic wave comprising of two mutually perpendicular waves, electric and magnetic.
EM Spectrum 10^-15 to 10^3m from shortest to longest in order:
Gamma, X-Ray, UV, IR,Visible, Microwave, S/M/L Radio
Visible Light 400-700nm
Speed Of Light 300,000,000 m/s (3x10^8) m/s
Time Period The inverse of wave frequency, 1s = 1Hz 0.001s =1000Hz 0.000,0003s =30,000,000s
V=Fλ Velocity = Frequency * Wavelength

Kilohertz 1,000 Hz = 1kHz (10^3)
Megahertz 1,000,000 Hz = 1MHz (10^6)
Gigahertz 1,000,000,000 Hz = 1GHz (10^9) then peta, exa, zetta, yotta increasing at 10^3

Millisecond 0.001s (ms) 10^-3s
Microsecond 0.000, 001s (μs) 10^-6s
Nanosecond 0.000, 000, 001 (ns) 10^-9s - then increasing in powers of (10^-3) pico, femto, atto, zepto, yocto
LO2 Factors Affecting Performance

Thermal Noise Random movement of electrons in a conductor, power is proportional to the temperature of the conductor. (Also called quantum noise)
Impulse Noise Spikes or disturbances (transient noise) more damaging to digital transmissions.
Cross Talk Noise Two conductors are placed two close together and the magnetic field from one couples into the other, causing a small ghost image to appear.
Near End NEXT pairs overly untwisted at the near end (low negative dB’s mean more noise), minimize with connectors properly fitted and tightly twisting wires.
Far End Less noise than near end FEXT – noise at the far end of the cable
PSNEXT Power Sum the cumulative effect of next from all wires in the cable mostly effects 1000BASE-T as its full duplex and so uses all the pairs.
ACR Attenuation -crosstalk ratio (dB) - difference between signal attenuation and near end cross talk.
SNR Signal:noise ratio +10dB means the signal has ten times, power of noise, +3dB means twice.
Attenuation -3dB means half of the power is lost, the opposite of amplifications, a form of distortion

LO3a Modulation 'modifying a carrier signal to carry information'

Audio Carrier Frequency of an audio carrier wave (F=1/T)
Amplitude Modulation Signal is contained in variations of the amplitude, simple but susceptible to noise and fading.
Modulation Index Ratio of signal to carrier – always less than 1
On off keying OOK – e.g. morse code is a form of AM-OOK
Frequency Modulation The frequency of the carrier is modulated (88-180MHz)
Phase Modulation The phase of the carrier is modulated e.g. 180degrees or 90degrees
Shift Keying A/P/F SK – Digital methods of modulation
QAM Quadrature amplitude modulation – ASK + PSK combined

Modulation Units Bits/Baud Baud Rate Bit Rate
ASK,FSK, 2-PSK Bit 1 N N
8-PSK, 8-QAM Tribit 3 N 3N
16-QAM Qaudbit 4 N 4N
64-QAM Hextabit 6 N 6N
LO3b Modems 'modulator demodulator'

V.21 Modem Uses FSK 1180HZ =0 980Hz = 1
V.34 Modem Baud 2400 – 12bits per signal change, 28Kbps
V.90 Modem PSK 24bits of data per baud 56Kbps
Baud Rate Signalling rate is number of signal units per second.
Data Rate Bit Rate, number of bits per second

Analogy: baud is analogous to a car while bit is the passengers (1 male, 0 female); the number of cars determines the traffic; the number of passengers does not.

LO4 Multiplexing 'the transmission of multiple pieces of data over the media
simultaneously'

Multiplexer MUX/DEMUX
Synchronous TDM Every input gets a time slot (wasteful with empty slots)
Asynchronous TDM Any input can have a vacant time slot (less wasteful)
Orthogonal FDM High speed transmission (wireless and digital TV)
WDM (Dense) Used with fibre in light, different channels have one of 80 wavelengths. 2.5Gbps (Sonet, Broadband, T1, E1)

L05a Transmission Modes

Baseband Signalling Uses the entire media bandwidth, using a change in DC on the line (Ethernet & RS-232-C)
Broadband Signalling Multiple simultaneous channels (cable TV or ADSL)
PSTN Bandwidth 300-3300Hz

Serial Transmission One bit after another, slower but only needs 1 connection (modem, network)
Parallel Transmission Multiple bits at the same time, not practical over distance (printers) max length 15-3ft before parallel Data Skew occurs which is when data arrives at RX at different times.

Asynchronous Communication 8,N,1 – Data, Parity Even or Odd, Stop Bits 1 or 2,
Asynchronous Overhead Simple to Implement but 25% bandwidth loss due to start/stop/parity.
ASCII American standard code for information interchange (7 bits represent 128 characters)
Parity Check Even parity or Odd parity
Unicode Supports up to 65, 000 language sets

Synchronous Communication Data is sent in large frames, start and stop marked with flags.
Synchronous Clocking Embedded in the data stream or on a clock line, RX and TX are always in sync, more efficient and can carry more data.
Ethernet Frame 64-1500bytes Frame Relay 4096 bytes
HDLC High level data link control
SDLC Synchronous data link control

L05b Satellite Communication

'increase the line of site needed for transmission beyond the curvature of the earth'

Minimum Escape Velocity 17,000mph for orbit
Period of Orbit Using kelpers theorem R3/T2 (radius *3 / period of orbit *3)
Low Earth Orbit 90 miles (LEO)
Geosynchronous Orbit 22,300 miles (24h orbit)
KU Band Low frequency microwaves 10.7-18GHz (Kurz-unten or directly below the K band)

L06 Digital Line Encoding

'the representation of digital information using discrete voltage signals – binary 1s & 0s are translated into sequences of voltage pulses that are sent along the media'

Unipolar Encoding Single voltage level, 1 is positive 0 is nothing, this is rarely used due to - DC component, a DC signal prevents the signal passing through transformers - Synchronisation as there is no clock.

Polar Encoding: Two voltage levels +ve and -ve are used to reduce DC component:

NRZ: Voltage is always either positive or negative never zero, this provides poor sync as there are long strings of the same bit.
NRZi: 1 is a change 0 is no change, good for strings of ones but bad for zeros.
RZ: Uses +/-/0 changes for each bit good sync, disadvantage is that 2 changes are needed for a single bit.

Biphase Encoding: (instead of returning to zero it goes to the opposite polarity)

Manchester: Changes in the middle of the bit interval up is a 1 change down is a 0 (used in Ethernet)
Differential Manchester: The voltage changes up and down to provide sync, zero is marked by a change at the beginning of the transition so there are two changes for a zero one for the data and one for the interval. (used in Token Ring)
Bipolar Encoding: Uses (z/+/-) positive is an alternation 0 is zero.

AMI: Alternate Mark Inversion 1 is positive and negative alternately, poor sync on long strings of 0s.
B8ZS: Bipolar 8-zero substitution. Violation occurs when 8 zeros occur based on the polarity of the previous one, this is used for American ISDN.
Previous 1 positive (000+-0-+) Previous one Negative (000-+0+-)

HDB3: High Density Bipolar3. Violation occurs if 4 zeros occur four violation triggers:

1) Number of 1s since the last substitution is odd, violation on the last zero.(000v)
2) Number of 1s since last substitution is even, violation on first and last zero. (v00v)
3) When polarity of the previous bit was positive both violations are negative. (+ |-00-)
4) When polarity of the previous bit was negative both violations are positive. (- |+00+)

all you need to know is, the polarity of each bit alternates- and the number of ones since the last substitution

4BT3: There is a look up table and there are ternary bauds (bandwidth +25% = block code)
2B1Q: Encodes 2 binary into 4 quaternary voltage levels (doubles bandwidth)
4D-PAM5: 1000Base-T, 9 voltage levels, 17 during transmission. (susceptible to noise due to all pairs being used)

Wednesday, 16 October 2013

Powers of 2 (Neckbeards Rejoice)

For varying levels of neckbeardage - the powers of 2 up to 2^32, handy for IPv4 subnetting:





01
12
24
38
416
532
664
7128
8256
9512
101,024
112,048
124,096
138,192
1416,384
1532,768
1665,536
17131,072
18262,144
19524,288
201,048,576
212,097,152
224,194,304
238,388,608
2416,777,216
2533,554,432
2667,108,864
27134,217,728
28268,435,456
29536,870,912
301,073,741,824
312,147,483,648
324,294,967,296































































VLSM & CIDR


The history of IPv4 Subnetting: VLSM & CIDR

A little guide I have written to support the information in CCNA Exploration semester 2 chapter 6 VLSM & CIDR, please let me know if you see any mistakes...

1969 – 1980: ARPANET addresses used only the first 8 bits, limiting the size of the whole system to 256 networks.

1981: RFC 790+91 introduce classful IP addressing A,B & C

1989: 159,000 hosts on the internet by October of this year

1993: RFC 1517+19 CIDR was introduced (classless addressing)

1993: RFC 1519 Class B addresses were depleting and found to be wasteful

1994: RFC 1631 Network address translation introduced (NAT)

1996: RFC 1918 Private addressing introduced

Classful IP addressing:


Class A 0.0.0.0 127.255.255.255
Class B 128.0.0.0 191.255.255.255
Class C 192.0.0.0 223.255.255.255
Multicast (D) 224.0.0.0 239.255.255.255
Experimental (E) 240.0.0.0 255.255.255.255

Notice how the number of high order bits shown here in green increases by one for each class also notice how the zero moves, shown by the red colour.

A = 00000000.00000000.00000000.00000000 to 01111111.11111111.11111111.11111111
B = 10000000.00000000.00000000.00000000 to 10111111.11111111.11111111.11111111
C = 11000000.00000000.00000000.00000000 to 11011111.11111111.11111111.11111111
D = 11100000.00000000.00000000.00000000 to 11101111.11111111.11111111.11111111
C = 11110000.00000000.00000000.00000000 to 11111111.11111111.11111111.11111111


Classful Subnets:

The network proportion of the address is shown in green, with the hosts in red, the subnet mask is shown for each class with its slashed short form written aside, notice how the mask follows the same pattern as the classful addressings

Address Subnet Mask

A = 00000000.00000000.00000000.00000000 to 11111111.00000000.00000000.00000000 /8
B = 10000000.00000000.00000000.00000000 to 11111111.11111111.00000000.00000000 /16
C = 11000000.00000000.00000000.00000000 to 11111111.11111111.11111111.00000000 /24

Hosts & Networks:

The powers of 2 are shown after each of the number of hosts, number of hosts is calculated by raising the number of bits in the host portion of the address (red) and raising it to the power of two, then subtract two as one needs to be reserved for the network and one for broadcast. Number of networks is calculated by raising the part in green to the power of two, taking away one bit extra for each class.

Number of networks in A = 08- 1 = 7 2^7=128 networks
Number of networks in B = 16- 2 = 14 2^7=128 networks
Number of networks in C = 24- 3 = 21 2^7=128 networks



A = 00000000.00000000.00000000.00000000=128(2^7) networks with 16777214(2^24)hosts/network
B = 10000000.00000000.00000000.00000000=16384(2^14) networks with 65534(2^16) hosts/network
C = 11000000.00000000.00000000.00000000= 2097152 (2^21) networks with 256 (2^8) hosts/network


Classful Routing Protocols (RIPv1 & IGRP)

These protocols assume the subnet mask by examining the network portion of the address, if the receiving interface of a router is in the same major classful network as the outgoing interface, it will retain the subnet mask, until a boundary router is reached, then the classful mask is applied.

CIDR Classless Inter Domain Routing

Introduced by the IETF allowed routers to disregard the network portion class of the host and instead use the subnet mask to determine the portion of the address to use for the network. Prefix aggregation allows more efficient use of address space. CIDR uses VLSM to allocate IP addresses, prefix aggregation is also known as route summarisation, a summary route is also known as a supernet as it encompasses several subnets. Route summarisation is the process of advertising a contiguous set of addresses as a single address with a less specific shorter subnet mask. A supernet is always a route summary, if a router has more than one entry for a route it will use the route with the highest mask, this would invalidate a supernet as its wouldn’t be so super... if there was a static route installed with a more specific bit match installed the router would prefer that route. A route summary is not always a supernet.

Calculating a Supernet (In three steps) I always mess this up for some reason

Step one = write the networks out in binary
Step two = determine the leftmost significant bit
Step three= set the non significant bits to zero to determine the network address

Monday, 7 October 2013

IIS 7.5 Lots Of Boring Information


IIS Notes: IIS 7.5, a mind numbing look into Microsoft IIS7.5 and then a greater look at how it can be used to monitor and optimise performance---- had a college exam all about it and these are my notes, this green header is to cheer up the proceedings...

Hardware & Software


Software:
ISS ships With Server 2008 R2 (Not enabled by default, enabled by 'turn windows features on or off'): A web server is made up of a computer, a web server OS and a web server program.

Min Hardware Requirements for server 2008 R2:
1.4Ghz 512Mb ram 10GB HDD and a DVDROM, 800x600 super VGA or higher graphics.

Features of IIS 7.5

Common HTTP Services:
Static Content, HTTP Errors, HTTP redirection

Application Development:
ASP.NET .NET Extensibility & ASP, CGI

Health & Diagnostics:
HTTP Logging, Logging Tools

Security:
Basic Authentication, Windows Authentication, IP Security, URL Authorisation

Performance:
Static Content Compression, Dynamic HTTP compression

Web Management Tools:
ISS Management Console, ISS Management Scripts & Tools

FTP Publishing:
FTP Server, FTP Management Console

Multiple User Domains:
ISS allows supporting multiple websites on a single server which conserves resources, websites can be assigned one of three unique Ids, a host header name, IP address or a TCP port number.

Security Issues

IP Spoofing:
hiding your real IP address by creating fake packets, ISS protects against this with Ingress Filtering.

Secure Socket Layer SSL:
allows one user to communicate securely with a web server using certificates authentication (https)

Configuration

ISS Manager:
A control panel for ISS full integrated with Application Server, Active Directory & windows sharepoint, the GUI makes adding and removing modules simple.

Web Management Service (WMSvc):
This service allows remote administration of the web server via the internet.

Monitoring Performance(Real Time or Report Based)

Task Manager:
You can see the status of the programs that are running, end programs that have stopped responding
and assess the activity of running processes using as many as fifteen parameters. Graphs and data on CPU and memory usage can be seen here.

Task Manager (Network Tab/Network Monitor):
On a network you can see how the network is functioning, who is connected, what they are working on and you can send p[people a message.

System Monitor:
Lets you configure counters to watch resource usage over time, on hardware resources performance of the machine and other remote machines and the activity of system services. You can choose what type of data to monitor (performance objects, counters or object instances) and the source of that data, this can be presented in a printable graph, a histogram or a report. Some objects include memory usage and system service usage. Sampling allows you to select on-demand or automatic monitoring, with start and stop times.
Performance Logs & Alerts
Allows collection of data from local or remote computers, viewing of logged counter data using system monitor or export to spreadsheets or databases. It is similar to the system monitor and allows generation of sampled reports. The user can define multiple logs and alerts to run, each one with a specified configuration.
Web Application Stress Tool (WAS) – free MS download
Lets you simulate an environment of a number of connections to stress test your site.


Improving Performance

Bandwidth Throttling:
Restricts the total network in bites\s, system monitor can verify this – enable through [media services | bit-rate throttling | set grouping by file type]

Bandwidth Limiting:
Limit bandwidth to a certain site using the html tag <site> <limits> maxBandwidth=”65536” maxConnections=”1024” connectionTimeout “00:01:00” </limits> </site> - can also be done in C# or VB script.

HTTP Compression:
Either server wide or per-directory, allows for faster transmission time between IIS and compatible browsers. Compressing application response files is called dynamic, static compression can be enabled by - Double click the local computer | right click websites folder then properties | check box for compression static | enter the path to the document for compression | set folder size (default is 95Mb)

Disabling Global compression the making it directory specific:
adsutil set w3svc/filters/compression/parameters/HcDoStaticCompression false1
adsutil set w3svc/1/root/Home/StyleSheets/DoStaticCompression true2

HTTP Keep-alives (enabled by default):
Keep the connection open while the server sends multiple elements enable by- expand the local computer, expand the websites folder and click properties | on the web sites tab click the check box to enable.

Connection Time-outs (default for www is 120s)
Help reduce the loss of processing resources consumed by idle connections, enable by - expand the local computer, click the websites folder and click on properties | on the web-site tab set the time in the connection time-out box, verify that the enable http keep-alives box is checked.

Maximizing Data Throughput for Network Applications:
Override windows preference for the files system by- control panel | double click network connection and click properties | select the file and printer sharing for Microsoft networks and click properties | on the optimisation tab select maximise data throughput for network applications.

Enabling CPU Monitoring:
Shuts down worker processes that consume large amounts of CPU time. IIS must be in worker process isolation mode, CPU monitoring is only available for application pools and doesn’t apply to CGI applications, to enable- expand the local computer, the the application pools folder in question and go to properties | in the maximum CPU box use the up/down arrows to set the max percentage of CPU usage, if the application uses more then an error is generated in the windows events blog | set the minutes to refresh CPU usage | check the box to decide on the action to perform when the limit is reached | write an event to the system log (no action) or choose shutdown to shut the process down when the max is reached | apply the settings.

Load Balancing/Mirrored Sites:
Adding more servers/adding a mirror to reduce load

Tuning the Server for TCP Optimisation:
TCP Slow start is a feature of the TCP/IP stack designed to reduce congestion using an algorithm, however this can cause performance to slow. This can be disabled using:
(netsh interface tcp set global autotuning=disabled).TCP listen intervals can be adjusted and TCP retransmission delay can be increased or decreased. 

------------------------------------


IIS 7.5 Summary – The absolute summary of this whole document is here:

Software 7.5 comes with Server 2008 R2
Min Req. 1.4GHzCPU, 512RAM, 10GbHDD, 800x600supervga, keys, mouse DVD

Features:
Static Content, HTTP errors, HTTP redirection, ASP, ASP.NET, .NET extensibility, HTTP Logging, Logging Tools, Basic Authentication, Windows Authentication, IP Security, URL Authorisation, Static Content Compression, Dynamic HTTP compression, ISS Management console, ISS Management Scripts & Tools, FTP Publishing Service, FTP Management Console, Multiple User Domains via IP, TCP or host header.

Security Issues:
IP Spoofing – Use Ingress Filtering, SSL certificates via HTTPS

Configuration:
Directly via management console or remotely via WMSvc web management services

Monitoring Performance:
Task manager, Task manager (network tab), System Monitor, Performance Logs & Alerts, WAS Web Application Stress Tool.

Improving performance:
Bandwidth throttling, Bandwidth limiting, HTTP compression, HTTP keep-alives (enabled by default) Connection time-outs (default 120s), Maximising data throughput for network applications, Enable CPU Monitoring, Load Balancing/Mirrored Sites, Tuning the Server for TCP Optimisation


  
Copyright (C)  2013  Michael Ferrie. Permission is granted to copy, 
distribute and/or modify this document under the terms of the GNU Free Documentation License, 
Version 1.3 or any later version published by the Free Software Foundation; 
with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. 
A copy of the license is included in the section entitled "GNU Free Documentation License"

Monday, 30 September 2013

Triple Boot Windows 7 Windows Server 2008 R2 & Debian 7 x64 on an HP ProBook6360b

So you need windows, windows server and Linux to run on a laptop and you don’t want to use visualisation because its mainly for girls. Here is how I did it...

Purchased an HP laptop from eBay, craching deal, came with windows 7 professional installed, and also some russian language toolbar in IE which leads  to have an air of suspicion about the device. For college I have to run win 7 and windows server 2008, I am also going to install Debian GNU/Linux running Gnome. The first thing you have to do is to shrink your NTFS partition.


HP ProBook, win7 pro x64 pre-installed - triple boot instructions, as a rule of thumb always do windows 1st.

1)Download GParted - get the live CD build.


2)Install tuxboot, if your in the debian userland including the *buntu, mint, etc etc just apt-get install tuxboot. Or get the package from here and build it. If you are running a windows os at this stage then I would recommend using the LILI usb creator for this, its super easy to use.

3) Get a usb stick and make it bootable using the usb creator you have downloaded, LILI shouldn't give you any trouble but if tuxboot tells you the device isn’t mounted then go into the command line and manually mount it using:

sudo fdisk -l
 
sudo mkdir /media/external
 
sudo mount -t vfat /dev/***sdb1*** /media/external -o uid=1000,gid=1000,utf8,dmask=027,fmask=137 

  
 ----/dev/***sdb1*** is the location of the usb stick, it could be sda or sdb2 etc etc 
 
 
4)Now switch off the machine and insert the bootable USB, boot into gparted, choose your language and type startx. Using the slider, shrink the NTFS volume down to the size you want, this laptop has a 250GB HDD so I decided to give the space out as follows 90 to windows 7, 90 to debian and 50 to server 2008. This process may take a while so be patient, mines took 11minutes.

5) Now shut down GParted and restart windows, checkdisk should appear, let windows run a check and restart it a couple of times to make sure you haven’t broken anything.

6) Install server 2008, I have server already burnt to a DVD so put it in and boot from the DVD, choose custom install and point the installation to the partition you made for it earlier with Gparted, I got it from dreamspark.

7) In the windows boot manager you should now have server 2008 and windows 7 installed, start and restart each of them to ensure they are working.

8) Download the latest Debian netinstall iso, should be about 280MB from the Debian website, then install Unetbootin (just use LILI again if your on windows). Grab a USB stick and make a bootable Debian disk, restart your PC and choose to boot from the USB disk.

9) When you install Debian you can save GRUB to the MBR and this will allow you to select your operating system.