If you have root privileges on your linux machine, you can create a symbolic link /Users that points to /home on your Linux machine. That shouldn't require changing anything.
Sent from my iPod On Feb 3, 2015, at 9:43 PM, "Li, Zhongwei" <[email protected]> wrote: > Dear CloudSuite Admins: > > Hi, during my last few days experiencing your benchmark suites, I found > something that may draw your attention as part of following your tutorial on > Web Search here. > > In this tutorial it assumes users may deploy your benchmarks in pure Linux > environments, yet in my case, I couldn't find a way to adapt it to my mixed > OS environment: > I'm trying to start the master on my Mac OS X (Yosemite) while run slave on > another machine with RedHat Linux distribution. Since they have a little > different directory structure -- for linux, it is /home/username... while for > Mac OS, it is /Users/username, the tutorial eventually lead me to the > following error msg when I was trying to start the cluster: > > starting namenode, logging to > /Users/username/nutch-test/search/logs/hadoop-ruby_-namenode-Gem-Garden.uta.edu.out > > xxx@slave1: bash: line 0: cd: /Users/username/nutch-test/search/bin/..: No > such file or directory > > xxx@slave1: bash: /Users/ruby_/nutch-test/search/bin/hadoop-daemon.sh: No > such file or directory > > localhost: starting datanode, logging to > /Users/username/nutch-test/search/logs/hadoop-ruby_-datanode-Gem-Garden.uta.edu.out > > xxx@slave1: bash: line 0: cd: /Users/username/nutch-test/search/bin/..: No > such file or directory > > xxx@slave1: bash: /Users/username/nutch-test/search/bin/hadoop-daemon.sh: No > such file or directory > > localhost: starting secondarynamenode, logging to > /Users/ruby_/nutch-test/search/logs/hadoop-ruby_-secondarynamenode-Gem-Garden.uta.edu.out > > starting jobtracker, logging to > /Users/username/nutch-test/search/logs/hadoop-ruby_-jobtracker-Gem-Garden.uta.edu.out > > xxx@slave1: bash: line 0: cd: /Users/username/nutch-test/search/bin/..: No > such file or directory > > xxx@slave1: bash: /Users/username/nutch-test/search/bin/hadoop-daemon.sh: No > such file or directory > > localhost: starting tasktracker, logging to > /Users/username/nutch-test/search/logs/hadoop-ruby_-tasktracker-Gem-Garden.uta.edu.out > > > Apparently the program is trying to use absolute path on my Mac OS (master) > machine to start daemon on the linux (slave) machine, however, I couldn't > locate where the relative code is in your benchmark. Or, how do I fix this. > > Please help and thank you in advance! > > Best, > Luke
