Freitag, 29. Oktober 2010
I was asked by the HPI administrators to warn them when we want to do large-scale crawling runs, so that they are prepared for the eventual traffic spikes. Since the lab machine is used by all seminar participants, it might be anyway a good idea to coordinate high load peaks by any of the sub-projects. Therefore, please give the docents a hint if you want to perform larger test runs.
Eingestellt von Peter Tröger um 04:58
Dienstag, 26. Oktober 2010
This is an initial overview of the lab setup being made available for the participants of this seminar. The resources are provided by the HPI FutureSOC Lab infrastructure.
Several aspects of the intended setup support the research on scalability issues. There are two different interconnection schemes to the stable storage, in order to analyze the implications of direct device access vs. networking file systems vs. iSCSI access.The internal connectivity is maximized, for both the networked file system access and the outer world connectivity. We also tried to determine some basic numbers for the available bandwidth. It turns out that the crawling efforts might be limited by the external connectivity of Campus Griebnitzsee to the Internet.
We will discuss about this picture in the seminar.
Eingestellt von Peter Tröger um 06:52
Mittwoch, 20. Oktober 2010
Donnerstag, 7. Oktober 2010
A group of hyper-intelligent pan-dimensional beings demand to learn the Ultimate Answer to the Ultimate Question of Life, The Universe, and Everything from the supercomputer, Deep Thought, specially built for this purpose. It takes Deep Thought 7½ million years to compute and check the answer, which turns out to be 42. Unfortunately, The Ultimate Question itself is unknown.