September 24, 2011 § Leave a comment

Cosmic Webin the late 70s, the start of the universe was described with the hot big bang theory; starting from a singularity and expanding, being filled by baryonic matter as well as non-baryonic matter. this model has successfully explained the existence of the cosmic microwave background and the amount of light being generated by nucleosynthesis. however, it had fundamental problems that it was unable to answer: the horizon problem (the homogeneity and isotropy of the universe in large-scale), the flatness problem (small curvature of the universe), the monopole problem (missing heavy stable magnetic monopoles), and the structure formation problem (the lack of support for the formation of the structures we can see today in the sky). in early 80s, research showed that the universe inflation could explain the above-mentioned problems even though there was still no concrete proof to support it [1].
among the different inflation theories proposed in the late 70s and early 80s, “old inflation” theory proposed by guth was the most popular, and easiest to understand, and was standardized as a textbook theory of inflation. however, this theory has shortcomings in the sense that it results in an inhomogeneous universe at the end. this led to the invention of the “new inflation” theory and later chaotic inflation scenario [2]. the inflation theory was supported by many practical data. after hubble space telescope was launched in the late 90s, the observation of distant supernovae became possible. it showed that distant supernovae appear to be fainter than expected in a flat matter dominated universe, which means at fixed redshift, they are at large distances than expected in such a context and thus the universe is accelerating [3]. another important proof is based on the observation and studies of the cosmic microwave background, which determines the total energy budget of the universe and proves that there is a shortage of about 70%, which leads to a missing exotic form of energy with negative pressure [4].  dynamics of our universe is defined by einstein’s equations, which shows that the contribution of energy content of universe that is represented by energy momentum tensor is related to the geometry given by the curvature of space-time. there are two ways to expand this equation to take into account the acceleration; either by supplementing an extra energy momentum tensor or by modifying the geometry itself. first method adds an extra momentum tensor, which is sourced by energy with negative pressure. the simplest candidate is the cosmological constant. however, due to its being non-evolving, cosmological constant theory has fine tuning implications. hence, dynamically evolving scalar field models such as phantoms, k-essence, tachyonic scalar fields and quintessence are introduced [5].
“the quintessence models were proposed to circumvent the two fundamental problems of the cosmological constant: a) a value of ΩΛ~0.7 is about 122 orders of magnitude from the naïve theoretical calculation b) it seems somewhat unnatural that we happen to live in a time when  ≈ 2 since this ratio depends on the third power of the redshift. …in “quintessence” models, the “dark energy” density tracks the development of the leading energy term making both comparable” [goobar 53], this modifies the late time evolution of the expansion rate of the universe, and explains the luminosity of supernovae and the angular distance of cosmic microwave background patterns. quintessence theory has its own shortcomings; the main problem lies in the fact that the quintessence field must be weakly coupled to ordinary matter. due to the coupling to ordinary matter, the quintessence potential does not decrease to zero in infinity as theory predicts [3].
looking at the present status of the studies and research on dark energy, it is fair to say that due to its being around for about 30 years, the dark energy could be a viable answer to the acceleration of the universe. however, its nature is still a mystery, and even though theories such as quintessence and tachyons are very successful in approaching it, it is still one of the biggest mysteries of cosmology.


  1. riazuelo, alain. “chapter 7, an introductory overview about cosmological inflation.”frontiers of cosmology proceedings of the nato asi on the frontiers of cosmology, cargèse, france from 8 – 20 september 2003. new york: springer, 2005. 101-138. print.
  2. linde, a.. “course 7. inflation and creation of matter in the universe.” the primordial universe l’univers primordial : les houches, session lxxi, 28 june-23 july 1999, ecole de physique des houches, ujf & inpg, grenoble. berlin: springer ;, 2000. 340-396. print.
  3. binétruy, P.. “Course 8. Cosmological Constant vs. Quintessence.” The primordial Universe L’univers primordial : Les Houches, Session LXXI, 28 June-23 July 1999, Ecole de physique des houches, ujf & inpg, grenoble. berlin: springer ;, 2000. 397-422. print.
  4. riazuelo, alain. “chapter 7, an introductory overview about cosmological inflation.” the invisible universe dark matter and dark energy. berlin: springer, 2007. 219-256. print.
  5. sami, m.. “chapter 8, models of dark energy.” the invisible universe dark matter and dark energy. berlin: springer, 2007. 219-256. print.
  6. goobar, ariel. “supernovae and dark energy.” particle physics and the universe proceedings of the 9th adriatic meeting, sept. 2003, dubrovnik. berlin: springer, 2005. 47-57. print.

trim your ssd

September 16, 2011 § 1 Comment

if you are building a system that uses solid state drive (ssd) which is a must if you are making industrial equipment that needs storage, make sure the operating system you use supports trimming. now that the solid state drives are widely used because of their price being comparable to the traditional drives’, operating system companies are adding the trim command support to their oses.

the trim command is introduced due to the difference in writing and over-writing performances on the solid state drives as opposed to the traditional moving parts drives that treat over-write as a write function. mlc nand flash based ssd’s (the common ssd drives in the market) smallest unit of storage is called a page which is 4kb. 128 pages comprise a block which is the ssd’s smallest erasable unit. this means that even though you are removing a small file, the whole block will not be erased if there are used pages in that block. the traditional way of removing files in oses is just by marking the page as not-in-use in os side, so the drive will not have any idea which pages are in use and which ones are not. these two facts will result in the drive not be able to wipe the blocks that are completely not-in-use. for example when you want to modify a small file on an ssd, the file will be marked not-in-use in os (traditional delete process) and then it will be written on a different block because the drive doesn’t know the current block can be wiped. this process is fine until the drive gets to a point that it goes out of empty blocks and needs to write to the already marked-as-delete pages. now the whole block has to be read into the memory then be erased from the ssd, the new data has to be inserted into the block and the block to be written to the disk again. this extra overhead is what kills the performance of the drives over long term.

trim is a command that lets the drive know which blocks are not in use and ready for garbage collection, this way the erase can be done when you delete a file rather than when you want to write on that block which result in avoiding the extra overhead.

linux started supporting the trim from version 2.6.33 onward, and windows with server 2008 server r2. for a complete list of the supporting oses see the wiki like here.

to add the support on linux you also need ext4 file system, for how to add trim support on linux please see this link.

inventing a story or telling lies?

September 5, 2011 § Leave a comment

life if pi by yann martedi finished the life of pi by yann martel a month ago, every time i decided to write something about it, i got stuck in the ‘so, what was it really about?’ state and procrastinate. i still don’t know how to interpret the book but i try writing something anyway.

i believe you can read the book hundred times and each time get a new idea. however, the book is very hard to read even twice because of its lengthy descriptive passages and in my opinion unrelated stuffs such as list of survival tips on a lifeboat, detailed zoological references, etc. (to be honest, i skipped a few of these overly emphasized paragraphs here and there) although i think those were there to fill the scientific portion of the book in order to show balance between its scientific and spiritual perspectives.

anyways, the book is about an indian boy who is interested in the core of most religions, god, and tries different paths of different religions to approach it. he is a son of a zoo keeper and has a lot of zoological hands-on experiences. past the few introductory first chapters, the story takes its major turn when he and his family emigrate to canada with a japanese ship. the ship sank and the boy is allegedly the sole survivor of the incident when he reaches to mexico after more than 200 days in a lifeboat with a tiger!

throughout this seem-to-be-simple story of survival, the reader goes through a lot of happenings. the story is told in first person form, by an adult person who is telling his life story using his childhood memories, which are now mixed with his adulthood believes (confusing, isn’t it?). underneath the simple story is a vast ocean of ideological, philosophical, spiritual and psychological hints that come in riddle form and need a puzzle solving mind to unravel them. reality and imaginations are so intertwined that the reader has to decide one by one as to which is which.

the best part of the book in my opinion is towards the end of the book, in the later chapters when pi is explaining his story to japanese maritime minister. the reader is in for a surprise, a huge turn that makes you question your whole understanding of the book, makes you review or even read the book again, just to be at peace with your mind. it has a deep philosophical essence which i relate it to hermeneutics. something very close to “big fish” or “the fall” movies, as the writer says, it’s all about how you like  to hear a story with animals better than one without. the dilemma of believing your past days being your best days or feeling bad about every minute of your life. positive versus negative perspective. you want to call it lying, you want to call it inventing, the writer’s idea is:

isn’t telling about something – using words – already something of an invention?

how to convert ext3 to ext4 on debian

September 4, 2011 § 1 Comment

recently i wanted to migrate from ext3 to ext4 file system without losing any data on the hard drive, in order to use the new trim functionality that is available on linux 2.6.33 and later. i searched the internet and there were a few links, however they were not complete or they were for different distros. i tried a few of them and ended up with not-bootable linux and lots of issues. finally i found how to do this by trial and error and thought to write a short tutorial on how to do it.

the first thing needs to be done is to enable the extra options on the ext3 file system by running the following command:

tune2fs -O extents,uninit_bg,dir_index [hard drive mount point]

in my case the mount point is /dev/sda1

then you have to run the file system check utility to re-adjust the file system to your new parameters (don’t worry about the errors, they are expected)

fsck -f -y [hard drive mount point]

after the file system parameters are adjusted you need to let the kernel know that you want to run ext4 file system. this is done through the grub boot loader. add the following kernel option to your kernel line in grub’s menu.lst file.


for example, your menu.lst section will be something like,

title Debian GNU/Linux, kernel 3.0.1-rt11
root (hd0,0)
kernel /boot/vmlinuz-3.0.1-rt11 root=/dev/sda1 ro quiet rootfstype=ext4
initrd /boot/initrd.img-3.0.1-rt11

then you need to edit /etc/fstab file and change the drive fs type to ext4, for example if the original line is

/dev/sda1 / ext3 rw,noatime,errors=remount-ro 0 1

you need to modify it to,

/dev/sda1 / ext4 discard,rw,noatime,errors=remount-ro,commit=0 0 1

discard option is added to use the new trim functionality for ssd drives. i will discuss this in another post.

now you can reboot and your system will boot up using your new ext4 file system. remember this is a must if you are going to use ssd drives on your system.

also note that grub doesn’t read ext4 file systems, you need to upgrade to grub2 for you boot loader to work. the upgrade to grub2 is very easy and harmless. here’s a good link on grub2

Where Am I?

You are currently viewing the archives for September, 2011 at artang.