Finding Good Quotes

When you grow

Surfing analogy for startups

There was an interesting talk a few days back in USC by Michael Sheha (his bio at the end of the post).

One of the analogies he gave was how startups were like surfing waves. The market is the wave and one needs to choose a big wave, a really big one. Otherwise there is no fun and no revenue. And the wave is something you can’t control. Its simply there and will come gushing. You cannot try to control it or tell it to move in another direction. You will have to ride along and ride skillfully.

Another analogy he gave was that a big company is like a warship or a tanker, they cannot change direction immediately, whereas you as a startup are like a PT boat - you are nimble, agile, fast and can change direction immediately (if you are not then you are doomed, pretty much).

One of the proverbs that really caught my attention:

The road to someday leads to the town of nowhere.

Here are other things that he talked about: * Customers really provide feedback when they PAY for something. * Keep costs low, really really low. You will never know when you will need the money. You should not repent that you could have saved earlier now that your funding got delayed. * Message matters - refine it a thousand times.

Here is his bio:

Michael Sheha is currently in-between startups. Previously he co-founded Networks In Motion (NIM) in 2000, which provided Location Based Services (LBS) to the world’s wireless carriers as a private labeled application and cloud-based solution, such as wireless personal navigation for mobile devices. NIM was started in 2000 and grew to over 300 employees internationally and to over $75m/yr in revenue. In 2009 NIM received the Southern California LAVA Award for the Best Exit in Internet & Technology when it was acquired by a public company. Prior to co-founding NIM, he worked at the California Institute of Technology Jet Propulsion Laboratory’s wireless communication systems and research section responsible for the design and development of digital and radio frequency communication systems, military GPS tracking systems, and the R&D in communication link and propagation studies. Michael graduated from University of Southern California (USC) with a MSEE in 2000, and Rensselaer Polytechnic Institute (RPI) with a BSEE in 1995. He is currently living in Southern California, CA with his wife and four children.

via MIND. IS BLOWN http://mindisblown.com/blog/2012/10/13/surfing-analogy-for-startups/

liveblogging osdi 2012 tuesday

There are four main sessions today. Looking forward to the Google talk on Spanner.

Distributed Systems and Networking

DJoin: Differentially Private Join Queries over Distributed Databases

Lots of data accumulated everywhere - social networks, hospitals, airlines..

Idea 1: give all data to a trusted party - but this may not exist. Idea 2: use secure multiparty computation but may talk long Idea 3: Use Differential Privacy.

Security

Improving Integer Security for Systems with KINT

Because the integers in C don’t have unlimited precision, it is possible for them overflow. For example 2³⁰ * 2² = 0. This can be exploited by attackers. One of the famous examples is that of the iPhone jailbreak. Another is the example of logical bugs in linux kernel. There is an OOM killer which assigns scores to processes based on memory usage and then kills the processes with the highest score. This can be exploited by malicious code that can take a lot of memory but still not be detected (because by overflow their scores can get evaluated to 0).

It is in fact hard to prevent integer overflows, even if you have unlimited precision (there could be other bugs or it is difficult).

Contributions of KINT:

• a case study of 114 bugs in the linux kernel


• KINT: a static analysis tool for C programs used to find the 114 bugs.

Case study: Linux kernel. The 114 bugs have been confirmed and fixed by developers. Most are memory and logical bugs.

Writing correct checks is non-trivial.

KINT has the following modules:

• Bound check insertion


• Taint analysis


• Range analysis

Advocates the use of NaN (instead of 0 when overflow occurs).

Details at http://pdos.csail.mit.edu/kint/

Dissent in Numbers: Making Strong Anonymity Scale

via MIND. IS BLOWN http://mindisblown.com/blog/2012/10/09/liveblogging-osdi-2012-tuesday/

Liveblogging OSDI2012

I am attending OSDI 2012 here at Hollywood, CA. Lots of interesting papers here and I will try to blog about this event. In particular I am excited about attending Google’s spanner talk scheduled for tomorrow afternoon (Tuesday).

The day didn’t begin too well, because I happened to witness a roadside accident. I was on the bus going to the Loews hotel (where the conference is going on), and the bus was waiting on red. It turned green and even before the bus moved ahead, a white Toyota Prius sped to turn left. Another car came dashing on the right of the bus lane because clearly it was green for it and before anybody noticed, there was a boom and a woman shouting - the Prius was hit on its right passenger side door. From what I figured out there was no injury of anybody. Other people were busy and my bus moved on. While this was a stupid accident that could have been avoided, I wish Vehicular Networks were mainstream now. If the Prius had alerted the driver about a car coming towards it, hopefully it wouldn’t have turned left prematurely. But more than vehicular networks, I wish

Keynote

The keynote is on cancer genomics. The speaker is David Haussler from UCSC. Here is the abstract:

Cancer is a complex condition—patients present with thousands of subtypes involving different combinations of DNA mutations. Understanding cancer will require aggregating DNA data from many thousands of cancer genomes, facilitating the statistical power to distinguish patterns in the mutations. The rapidly plummeting cost of DNA sequencing will soon make cancer genome sequencing a widespread clinical practice. To anticipate this, UCSC has built a 5-petabyte database for tumor genomes that will be sequenced through National Cancer Institute projects—the Cancer Genomics Hub—and is tackling the significant computational challenges posed by storing, serving, and interpreting cancer genomics data.

Some of the questions/points raised:

• there is an enormous opportunity to bring big data techniques to cancer genomics.


• how do we find out mutations from these gene data.


• how to map these mutations to the pathways that lead us to cancer, which should help us prevent these cancers.

Flat Datacenter Storage

• FDS is simple, scalable blob storage


• distributed metadata management


• Built on a CLOS network with distributed scheduling.


• High read/write performance


• fast failure recovery


• high application performance.

Data is organized as blobs, and each blob has multiple tracts.

Consists of: - Tractserver: sits between raw disk and network. - Metadataserver: - Client

GFS/Hadoop have the following problems: - Centralized metadata server - critical path of reads/writes - large (coarsely striped) writes

DHTs: - multiple hops to find data - slow recovery

FDS tries to position itself in between.

There is a tract location table, that maps for each locator the disks it has to read.

CLOS:

Generally we have this tree structure for the DC architecture. FDS provisions as much bandwidth as each disk requires. Full bisection bandwidth is only stochastic. Long flows are bad for load balancing. FDS generates a large number of short flows are going to diverse desitnations But TCP likes long flows. FDS creates “circuits” usign RTS/CTS.

via MIND. IS BLOWN http://mindisblown.com/blog/2012/10/08/liveblogging-osdi2012/