Tmux I sometimes need to compile software package in the High Performance Computer(HPC). The compiling process often takes very long because C++ never pays any attention to this (Thank google, we now have golang that has a significantly shorter compiling time). It will become a serious trouble for remote compiling, because network could go wrong at any moment therefore, the compiling process in the HPC will be interrupted. Unfortunately, remote access comprises majority of the working senario: we used a ssh client (say, putty or iTerm) to connect to the HPC and excute command remotely.

As one of my daily works, checking files from simulation results is of course my practical need. Therefore, I am always interested in using different ways to do it. One of my previous post, “checking_files_in_subfolders_using_power_shell” was on how to use power shell in windows 10 to do this job. Here I am going to test other ways that I used even before powershell script. 1 find This is a Linux command and it is actually my first way to iterate folders to find files, since the operation system I work with in Supercomputer is Linux.

As daily work, sometime I need to check some log files from certain calculations – a large number of calculations. Usually they contain some keyword or sentence, like “Normal termination” or “Calculation ended” or “succussefully conveged” for me to search. This is a typical job that can be done rountinely in linux using commands like “grep”, “awk”. However, after download those file to a windows PC, it would be more convinient to use windows shell to do, although there exists some “tranplanted” version of some of the linux shell commands that are complied with native win32 toolchains.

Many faces of the writing Writing can be for many purposes and they are all different. For a story writing or an anecdote, you have many adjectives for different things. For analytical essay, you have many judgements and reasoning words to piece up the logic. And for academic journal writing, you want to have plain language and fluent logic. ​ It is very interesting that in academic writing, you both want to avoid jargon and use jargon.

Octopus is a scientific simulation software for solving quantum problem. (http://www.tddft.org/programs/octopus/wiki/index.php/Main_Page) Recently, there is a new version 6.0 released (2016-09-05), which has a major improvement. Here we try to install it on the Centos 6.8 (with minimal installation) First, install Centos 6.8 with minimal installation choice. Installation of octopus 6 mainly follow their manual here : http://www.tddft.org/programs/octopus/wiki/index.php/Manual:Building_from_scratch 1. install several compilers to use. You can check your current available compilers by

Updated on 20170604: I put in more content to setup on tools chain for python development environment. Extra options were deleted. I bought a VPS from ramnode last time (last blog), so I guess it is a good time to deploy some web application on it. Since the python came with Centos 6.8 is kind of old (python 2.6), the first step I take is to upgrade the python compiler in Centos 6.

Virtual Private Server (VPS) is a new way to host a personal server. It hides the hardware detials from the user (which is safe) and delivers a flexible (sometime even real time adjustable) performance to the server. They usually have two types: KVM and OPENVZ. The difference of the two virtualization method is the the OPENVZ is container-like, which is also limited to linux only, but KVM use the main OS system to simulate the hardware for user-OS, which theoretically allows any type of the operation system. Because of the working principle of KVM, it could provide better isolation (Those guys doing virus analysis would like this). However, if you are looking for better performance of the Linux system, openvz maybe a better choice.

VPS from commercial providers is a perfect source to setup a personal server. This article will include:

• Bought a VPS

• Benchmarking the performance of the VPS

​