Hello Jianwu,
thanks for your suggestions, but I think this issue still remains. Maybe
we have slightly different use-cases.
Use-case:
- I want to start multiple instances of Kepler at the same machine
- I want each of Kepler instances to be executed in separated HOME
I am using following scripts to achieve that:
(both scripts are executed inside $HOME/kepler_temp)
--- CUT multiple.sh ---
#!/bin/tcsh
mkdir $1
cd $1
setenv OLD_HOME $HOME
setenv HOME `pwd`
$OLD_HOME/kepler/kepler.sh -D"user.home"="$HOME"
--- CUT ---
--- CUT start_test.sh ---
#!/bin/tcsh
./multiple.sh 0 &
./multiple.sh 1 &
./multiple.sh 2 &
--- CUT ---
- Create directory kepler_temp in your $HOME
- put multiple.sh and start_test.sh inside $HOME/kepler_temp
- start ./start_test.sh
As you can see, only one Kepler instance is started.
Michal
Hi Michal,
In Kepler trunk, if I run command like ./kepler.sh -D"arg1"="value1",
the workflow can get the new value. You can get its value in the
workflow by using property("arg1") in Expression actor. I also tried
./kepler.sh -D"user.home"="$HOME/test-home", which also works for me.
Please try whether this way works for you.
Your patch looks good. Just it is not general. Users may also want to
pass other properties. Luckily, the current kepler trunk support any
property to be set after kepler.sh. At least in my tests.
Best wishes
Sincerely yours
Jianwu Wang, Ph.D.
[email protected]
http://users.sdsc.edu/~jianwu/
--
Michal Owsiak <[email protected]>
Institute of Bioorganic Chemistry PAS
Poznan Supercomputing and Networking Center
Noskowskiego 12/14, 61-704 Poznan, POLAND
http://www.man.poznan.pl
_______________________________________________
Kepler-dev mailing list
[email protected]
http://lists.nceas.ucsb.edu/kepler/mailman/listinfo/kepler-dev
