Replicating a post from VirtualBox forum for self reference and archival. Original port is available here. Additionally, on my MBP, I could not see vboxnet0 adapter. The procedure to make it available is as follows:

1. Goto VirtualBox Preferences (⌘ + ,)
2. Click on 'Network'
3. Click on the "+" sign to add the adapter

You are now set to follow the post below.

----

Has anybody been able to communicate by SSH between a Mac OS X 10.6 (Snow Leopard) host and a guest running on VirtualBox? My guest is Ubuntu Server 10.10 (64-bit). I have done the following, but to no avail:

1. Installed OpenSSH Server on my Ubuntu Server 10.10 guest.
2. Shut the guest down, and used VirtualBox Settings to enable a second virtual network adapter, named (by default) vboxnet0, and which I set as a host-only adapter.
3. Ran ifconfig vboxnet0 in my host terminal:

   $ ifconfig vboxnet0
   vboxnet0: flags=8843 mtu 
   ether 0a:00:27:00:00:
   inet 192.168.56.1 netmask 0xffffff00 broadcast 192.168.56

4. Booted Ubuntu Server up again, and edited the /etc/network/interfaces file (sudo emacs /etc/network/interfaces) to append the following lines:

   auto eth1
   iface eth1 inet static
   address 192.168.56.10
   netmask 255.255.255.0

5. Saved and closed the interfaces file.
6. Ran sudo ifup eth1 from the Ubuntu command line.
7. Switched to my host terminal and entered ssh 192.168.56.10.

But my Mac OS X ssh client just hung there without a response until it timed out. :-( There must be something else I am supposed to do to make this work—but what could it be? Does anybody know? Thank you!

Permalink | Leave a comment  »

Permalink | Leave a comment  »

In my professional life, I use python on a day to day basis, but I solve the commercial world problems like how to make an application work better without flaws or how to make sure that a web-app works the way it is expected to work. I don't solve mathematical problems at work nor do I foresee any work in solving mathematical problems.

The only reason I went there was to vet my decision to move to using Python as my primary language for coding. About 3 years back, when I decided that I should have a skill set beyond MS Visual Basic / .Net and pick up an open source coding platform, I had a choice between Java and Python. Ruby and Rails were also promising and upcoming languages at that time. But I picked Python.

I am glad I picked up Python back then and after attending a scientific python conference I can safely say, python was the correct choice I made back then.

Permalink | Leave a comment  »

Permalink | Leave a comment  »

Permalink | Leave a comment  »

Here are my thoughts for you. Just like the way you decide, who to talk to by remaining invisible, I am going to decide whether to respond to your chat ping or not.

Once in a while u r in invisible mode, I can understand, but everyday? If you don't want other's to ping you as you are busy, set the busy status & politely tell ppl, you will get back to them. If you have a pesky friend on the list delete / block him or her.

Note: This post is not for any one person but in general to all my invisible friends.

Permalink | Leave a comment  »

Today morning, I was monitoring two EC2 AMI’s for their performance under load conditions for one applications in office. I was doing it in the most rudementary method, by obeserving the output from ‘top’. The last parameter in the CPU information line, ‘st’, drew my attention immediately as it was consuming nearly 60% of CPU time!

1:Def - 22:52:40 up 15:20,  1 user,  load average: 7.07, 3.30, 1.44
 Tasks: 110 total,   5 running, 105 sleeping,   0 stopped,   0 zombie
 Cpu(s): 30.9%us,  7.3%sy,  3.7%ni,  0.0%id,  0.0%wa,  0.0%hi, 0.0%si, 58.1%st
 Mem:   1747764k total,  1687264k used,    60500k free,    34288k buffers
 Swap:   917496k total,       88k used,   917408k free,   538928k cached

I guess I had missed it or never paid attention to that parameter earlier. Googled a little to find out that ‘st’ = ‘steal’, in the literal sense.

Steal time is the percentage of time a virtual CPU waits for a real CPU while the hypervisor is not scheduling this virtual CPU.

I found a very nice presentation from IBM, which explains what steal time is and why it exists –> http://t.co/XNhOIan

Permalink | Leave a comment  »

See the full gallery on Posterous

Anyways, here comes the Indian jugaad to rescue! I took one pin of a broken 3 pin adapter & plugged it into another 3 pin international converter adapter. This pin can be aligned to coincide with the earthing or grounding plate of the plug provided by Apple. See attached photos for better understanding.

Permalink | Leave a comment  »

Following up on the previous posts on configuring apache on Ubuntu, I now post the method to do the same on the Mac.

Step 1: You will need to download the mod_proxy_html module from http://apache.webthing.com/mod_proxy_html/

Step 2: Compile the code.

$ sudo apxs -ci -I /usr/include/libxml2 mod_xml2enc.c
$ sudo apxs -ci -I /usr/include/libxml2 -I . mod_proxy_html.c

Then update you httpd.conf to include

LoadFile /usr/lib/libxml2.dylib
LoadModule  proxy_html_module   libexec/apache2/mod_proxy_html.so
LoadModule  xml2enc_module   libexec/apache2/mod_xml2enc.so

Also take a look at the proxy-httpd.conf file that ships with mod_proxy_html for additional config you may want.

Permalink | Leave a comment  »

Permalink | Leave a comment  »

This summer like every summer I could not camp, I spent my days in an AC office, hammering away at the keyboard, just like any other 'IT Worker'. Last week I saw a small window to make good an escape from the AC office to a Tiger Sanctuary, before they closed down for the monsoons. Making a mad dash, I managed to join a group of friends heading to Kanha Tiger Reserve. The forest looked just plain beautiful, with shades of green and brown. Never had I had a chance to see Kanha in this season. It seemed like all the deers in the forest had come out into the meadows to celebrate and get wet in the rains. The male chital (spotted deers) sported beautiful antlers.

This July calendar, I send across a picture of one such male Chital.

Classic square monitor (4:3)
New modern full HD monitors, widescreen (16:9)
Laptop or HD compatible monitors, widescreen (16:10)

To download,
Firefox users: Right-click on the appropriate link and click “Save link as” to download.
Internet Explorer users: Right-click on the appropriate link “Save target as” to download.

Permalink | Leave a comment  »

Since Swype comes as system app, you cannot update it or remove it
using the regular method. To remove old swype, you need to first root
your phone and then delete the swype.apk and then install swype.

Procedure I followed:
1. Did the "Super One Click Root(2.2)", instructions are available on
xda-developers forum thread
2. Then using adb, logged into the phone, executed the following
commands to get rid of default swype
su mount -o rw,remount -t yaffs2 /dev/block/mtdblock3 /system rm -r /system/app/Swype* reboot
3. Registered for the beta program with swype

Permalink | Leave a comment  »

---------- Forwarded message ----------
From: Seth Godin
Date: 27 May 2011 16:01
Subject: Seth's Blog : Caring
To:

Caring

No organization cares about you. Organizations aren't capable of this.

Your bank, certainly, doesn't care. Neither does your HMO or even your car dealer. It's amazing to me that people are surprised to discover this fact.

People, on the other hand, are perfectly capable of caring. It's part of being a human. It's only when organizational demands and regulations get in the way that the caring fades.

If you want to build a caring organization, you need to fill it with caring people and then get out of their way. When your organization punishes people for caring, don't be surprised when people stop caring.

When you free your employees to act like people (as opposed to cogs in a profit-maximizing efficient machine) then the caring can't help but happen.

[You're getting this note because you subscribed to Seth Godin's blog.]

Permalink | Leave a comment  »

===========

Created by Cynick Young. Last edited by David Jones

One of the challenges in testing a GWT application using a tool like
Selenium, is to be able to accurately tell when all the asynchronous
requests have finished and that your application is in a state you
expect. I've searched around the internet and didn't find any useful
solutions on this topic so I'll share my implementation.

Defining Our Ajax Hook

Since GWT is all about Ajax, we need to inject some Ajax hooks into
our production web application. I'll create a new Semaphore that will
be included as part of the javascript that gets loaded into the html
of the GWT application. This will be our door in.

semaphore.js:
var IN_FLIGHT_AJAX = IN_FLIGHT_AJAX || function() {    // private variable bound through closure    var activeCallCount = 1;    return {        callsInFlight: function() {            return activeCallCount;        },        decrementCallsInFlight: function() {            activeCallCount--;        },        incrementCallsInFlight: function() {            activeCallCount++;        }    } }();

Notice how I've initialized the counter to be 1. This is important
because the GWT bootstrap process in itself can take a bit of time.
So, at a minimum we have to wait until the module is fully loaded.
Otherwise, you will get the occasional timing/synchronization failure
during test runs.

Bridge Between GWT and Our Testing Hook

I'll create a JavaScript Native Interface (JSNI) Java class to provide
the ability to manipulate our semaphore from within our GWT
application.

InFlightAjaxCallMarker.java
public class InflightAjaxCallMarker {    private boolean hasStartedCall = false;    private boolean hasEndedCall = false;    public InflightAjaxCallMarker() {        super();    }    public void startCall() {        if (!hasStartedCall) {            incrementCallsInFlight();            hasStartedCall = true;        }    }    public void endCall() {        if (!hasEndedCall) {            decrementCallsInFlight();            hasEndedCall = true;        }    }    private native void decrementCallsInFlight() /*-{        try {            if (($wnd.IN_FLIGHT_AJAX) && (typeof $wnd.IN_FLIGHT_AJAX.decrementCallsInFlight === "function")) {                $wnd.IN_FLIGHT_AJAX.decrementCallsInFlight();            }        } catch (ex) {            // Do error handling        }    }-*/;    private native void incrementCallsInFlight() /*-{        try {            if (($wnd.IN_FLIGHT_AJAX) && (typeof $wnd.IN_FLIGHT_AJAX.incrementCallsInFlight === "function")) {                $wnd.IN_FLIGHT_AJAX.incrementCallsInFlight();            }        } catch (ex) {            // Do error handling        }    }-*/; }

Marking Our Module Loaded

As I mentioned earlier, we want to wait for our module to be fully
loaded. This can be accomplished by adding a call to the end of
onModuleLoad().
public void onModuleLoad() {        ... perform some stuff        new InflightAjaxCallMarker().endCall();    }

Marking Our Ajax Calls

The easiest way to mark all our Ajax calls is to go through the
AsyncCallback interface. Since all our asynchronous interactions
require a callback, we can create a callback template to perform our
marking. From this point onward, developers will create new
AsyncCallbackAdapter classes instead of AsyncCallback.

AsyncCallbackAdapter.java
public abstract class AsyncCallbackAdapter<T> implements AsyncCallback<T> {    private InflightAjaxCallMarker inflightAjaxCallMarker;    public AsyncCallbackAdapter() {        this.inflightAjaxCallMarker = new InflightAjaxCallMarker();        this.inflightAjaxCallMarker.startCall();    }    @Override    public final void onFailure(Throwable caught) {        try {            onFailureBody(caught);        } finally {            inflightAjaxCallMarker.endCall();        }    }    protected abstract void onFailureBody(Throwable caught);    @Override    public final void onSuccess(T result) {        try {            onSuccessBody(result);        } finally {            inflightAjaxCallMarker.endCall();        }    }    protected abstract void onSuccessBody(T result); }

Getting Selenium to Wait

Finally, we will create a Selenium function to test our Ajax status.
String script = "selenium.isAjaxComplete = function() {    var appWindow = selenium.browserbot.getCurrentWindow();    if ((appWindow.IN_FLIGHT_AJAX) && (typeof appWindow.IN_FLIGHT_AJAX.callsInFlight === 'function')) {        return (appWindow.IN_FLIGHT_AJAX.callsInFlight() === 0);    }    return false; }"; selenium.addScript(script, scriptTagId);

Now, we can force Selenium to wait for Ajax calls to finish whenever we need.
selenium.waitForCondition("selenium.isAjaxComplete();", timeoutValue);

How Does the Semaphore Work With Multi-Threaded Rendering?

There are a few things that need to be clarified before I start
explaining about the workings of the semaphore.

The semaphore needs be included on each html page.
Javascript doesn't have threads. In all browser implementations, there
is only one thread of execution for javascript in each browser
window/tab.
Each browser will also have a native timer thread and any number of
native threads to do asynchronous XMLHttpRequest (XHR) calls. These
native threads are not accessible nor can we count on them in
javascript. When an asynchronouse response returns from the server, it
is the browser implementation that places the callback code onto the
javascript execution stack. No assumption should be made about the
order of execution.
In javascript, there are only 2 variable scopes: global, and function.

Now, on to the details.

Our semaphore is created in the global (top window) scope. If an
existing object already exists, it will use that existing object.
Otherwise, it will execute an anonymous function that returns us a new
object.
var IN_FLIGHT_AJAX = IN_FLIGHT_AJAX || function() {   ...    return {        ...    } }();

Our private counter is only accessible inside our semaphore object
because we defined it using function scope. Outside of the function,
activeCallCount cannot be modified and this validates the semaphore.
Through the "magic" of closure, we can provide accessors to our
counter even after the function has returned. Since javascript is
single threaded execution, we are guaranteed to serialize the
increment and decrement of the counter.
// private variable bound through closure    var activeCallCount = 1;
With multiple browser windows or tabs, there will be a single
semaphore loaded and associated within each window/tab. This works in
conjunction with the fact that each window/tab has its own global
environment. The asynchronous XHR calls in each window/tab will only
affect its associated semaphore, and that's what we want.

Conclusion

This is a comprehensive GWT testing solution for use with tools like
Selenium. The semaphore implementation can be used with any Javascript
framework, not just GWT. The JSNI bridge along with the AsyncCallback
template is the key to providing a seamless integration with GWT
applications. This approach to waiting for Ajax calls to finish will
deter developers from writing brittle web tests that depend on
specific pieces of text or UI elements. The drawback to this
implementation is that production code will contain small amounts of
testing hooks.

Permalink | Leave a comment  »

A funny email I received from an acquaintance, I really do not know how to respond to this email.
Should I talk some sense into him or just laugh it off!!

*****************************
Hi guys,

Permalink | Leave a comment  »

Following up on the last post, here I am outlining how to make reverse
proxy work with SSL,
$ sudo a2enmod ssl*
to enable all Apache SSL modules.
After this I appended the following code in the file
/etc/apache2/sites-available/default-ssl
AddDefaultCharset off Order deny,allow Allow from all SSLProxyEngine on SSLProxyCACertificateFile /etc/apache2/cacerts.crt ProxyPass /abc https://xxx.xxx.xx.xx:4400/abc ProxyPassReverse /abc https://xxx.xxx.xx.xx:4400/abc ProxyVia off
Since I was re-directing to a secure server, I faced two errors:
[error] [client xxx.xxx.xx.xx] SSL Proxy requested for 127.0.1.1:443 but not enabled [Hint: SSLProxyEngine] [error] proxy: HTTPS: failed to enable ssl support for xxx.xxx.xx.xx:4400 (xxx.xxx.xx.xx)
The main reasons being, first I had forgotten to add the
SSLProxyEngine parameter to the configuration.
Secondly, when apache acts as a proxy, it acts like a client
requesting over SSL. Hence it needs the server certificate (PEM
encoded) to do a successful handshake.
Using openSSL,
openssl s_client -connect xxx.xxx.xx.xx:4400 -showcerts
Copied the encoded content from -----BEGIN CERTIFICATE----- & -----END
CERTIFICATE----- , saved it to cacerts.crt.
After this apache started serving SSL access to remote server via proxy.

Permalink | Leave a comment  »

http://localhost/abc was mapped to http://remote_aws_ip:8888/abc

To do this, follow the steps below:
$ sudo apt-get install libapache2-mod-proxy-html $ sudo a2enmod proxy*

Create a new file, remote_aws_ip, inside /etc/apache2/sites-available
<VirtualHost 127.0.0.1:80> DocumentRoot "/var/www" ServerName remote_aws_ip <Proxy *> AddDefaultCharset off Order deny,allow Allow from all </Proxy> ProxyPass /abc http://remote_aws_ip:8888/hcm ProxyPassReverse /abc http://remote_aws_ip:8888.com/hcm ProxyVia off </VirtualHost>

Then enable the site, by running
$ sudo a2ensite remote_aws_ip
This will make a symlink to the file into sites-enabled.

Last but not the least
$ sudo service apache2 restart

Now I point my browser to http://localhost/abc and it fetches content
from http://remote_aws_ip:8888/abc

Permalink | Leave a comment  »

Faced the following error while initialising a virtualenv instance on ubuntu 11.04:

aditya@TR:~$ virtualenv virtualenvs/env_selenium --no-site-packages New python executable in virtualenvs/env_selenium/bin/python Installing distribute................................   Complete output from command /home/aditya/virtualenvs/env_s...ython -c "#!python \"\"\"Bootstrap distribu...1:]) " --always-copy -U distribute:   Extracting in /tmp/tmptp7iU4 Traceback (most recent call last):   File "<string>", line 477, in <module>   File "<string>", line 473, in main   File "<string>", line 74, in _install   File "<string>", line 444, in _extractall   File "/usr/lib/python2.7/tarfile.py", line 2083, in extract     self._extract_member(tarinfo, os.path.join(path, tarinfo.name))   File "/usr/lib/python2.7/tarfile.py", line 2159, in _extract_member     self.makefile(tarinfo, targetpath)   File "/usr/lib/python2.7/tarfile.py", line 2199, in makefile     copyfileobj(source, target)   File "/usr/lib/python2.7/tarfile.py", line 266, in copyfileobj     shutil.copyfileobj(src, dst)   File "/usr/lib/python2.7/shutil.py", line 48, in copyfileobj     buf = fsrc.read(length)   File "/usr/lib/python2.7/tarfile.py", line 817, in read     buf += self.fileobj.read(size - len(buf))   File "/usr/lib/python2.7/tarfile.py", line 735, in read     return self.readnormal(size)   File "/usr/lib/python2.7/tarfile.py", line 744, in readnormal     return self.fileobj.read(size)   File "/usr/lib/python2.7/gzip.py", line 252, in read     self._read(readsize)   File "/usr/lib/python2.7/gzip.py", line 299, in _read     self._read_eof()   File "/usr/lib/python2.7/gzip.py", line 338, in _read_eof     hex(self.crc))) IOError: CRC check failed 0xbf8917c9L != 0x59a4e93eL ---------------------------------------- ...Installing distribute...done. Traceback (most recent call last):   File "/usr/bin/virtualenv", line 3, in <module>     virtualenv.main()   File "/usr/lib/pymodules/python2.7/virtualenv.py", line 536, in main     use_distribute=options.use_distribute)   File "/usr/lib/pymodules/python2.7/virtualenv.py", line 626, in create_environment     install_distribute(py_executable, unzip=unzip_setuptools)   File "/usr/lib/pymodules/python2.7/virtualenv.py", line 367, in install_distribute     _install_req(py_executable, unzip, distribute=True)   File "/usr/lib/pymodules/python2.7/virtualenv.py", line 339, in _install_req     cwd=cwd)   File "/usr/lib/pymodules/python2.7/virtualenv.py", line 597, in call_subprocess     % (cmd_desc, proc.returncode)) OSError: Command /home/aditya/virtualenvs/env_s...ython -c "#!python \"\"\"Bootstrap distribu...1:]) " --always-copy -U distribute failed with error code 1

Hunting around the web realised the issue was with the replacement of setuptools with distribute. To get rid of the issue, updated virtualenv installation.

$ sudo pip install pip --upgrade $ sudo pip install virtualenvwrapper --upgrade

Permalink | Leave a comment  »

This May, I dig into the archive to select a photo I clicked last may in Kanha.

There is no need for me to describe or say much about the national bird of India.

I hope you like it and look forward to your comments & critiques.

To download please click on the links below.

Classic square monitor (4:3)
New modern full HD monitors, widescreen (16:9)
Laptop or HD compatible monitors, widescreen (16:10)

To download,
Firefox users: Right-click on the appropriate link and click “Save link as” to download.
Internet Explorer users: Right-click on the appropriate link “Save target as” to download.

Permalink | Leave a comment  »

The colour of the month is Blue, because India entered the ICC World Cup 2011 finals. As a token of my support to the Indian Cricket Team, I have chosen a blue coloured bird to adorn your desktops.

The bird in the image is the Tickell's Blue Flycatcher (Cyornis tickelliae)

This month, the image will not have the calendar on it. I tried putting it at various locations, but it was spoiling the entire image.

Also, I have uploaded only 3 files, with the following sizes:

1. 1920 x 1200 (16:10)
2. 1920 x 1080 (16:9)
3. 1024 x 768 (4:3)

These days a lot of new laptops and new screens come in HD resolutions, i.e. 16:9 ratio, older screen came in a 16:10 ratio. If your monitor is a newer full HD monitor you can download the file number 2, if your monitor is a little older, you can download file number 1. If your monitor is of the real old type, square screen, please download file number 3.

And in-case neither of them fits your screen, please let me know the screen resolution, I shall custom make one for you and send it across.

To download,
Firefox users: Right-click on the appropriate link and click “Save link as” to download.
Internet Explorer users: Right-click on the appropriate link “Save target as” to download.

Permalink | Leave a comment  »