Differences

This shows you the differences between two versions of the page.

--- cluster:106 [2012/07/10 18:30] – hmeij
+++ cluster:106 [2016/11/21 20:22] (current) – [Sample] hmeij07
@@ Line 10: / Line 10: @@
 ==== Sample ====
-* save in a script and submit to scheduler
+* save in a script say called ''run'' and submit to scheduler ''bsub < run''
 <code>
@@ Line 21: / Line 21: @@
 # find queues: bqueues
 # submit job: bsub < run
-# jobs submmitted: bjobs
+# show jobs submitted: bjobs
 #BSUB -q stata
@@ Line 30: / Line 30: @@
 # use n cores (job slots) ... a license limitation of 6
 #BSUB -n 6
-# force using all on one node (hosts=1) or use the -x flag
+# force using all on one node (hosts=1)
 #BSUB -R "span[hosts=1]"
-# can stata-mp work across nodes?
-# it appears not so (only one out of 3 hosts ran stata-mp).
-# if so, raise the hosts=value or control
-# nr of processes per node with tile=value
-####BSUB -R "span[ptile=2]"
+# can stata-mp work across nodes?
+# it appears not
 # scratch dirs
@@ Line 48: / Line 45: @@
 # alternate workdir, stage data and/or retrieve results
 # cd $MYSANSCRATCH
-# cp ~/stata/AllRegions2.dta  ~/stata/nikkei1.dta .
+# cp ~/stata/AllRegions2.dta  ~/stata/nikkei1.dta  .
 # /share/apps/stata/10/stata-mp -b do wagepan80_87
-# assumes output is written back to home, else copy back
+# assumes output is written back to home, else copy results to ~/stata/
 # working in home dir
+cd ~/stata/
 /share/apps/stata/12/stata-mp -b do firm_analysis_cluster