Browse files

tar wasn't compatible on Mac OS in add_tgz_to_build.sh

  • Loading branch information...
1 parent 195ea77 commit 6c5660b05d7b176a9132d6efd330099146e1195b Gabor Szabo committed Aug 1, 2012
Showing with 3 additions and 2 deletions.
  1. +1 −1 add_tgz_to_build.sh
  2. +2 −1 remote_deploy.sh
View
2 add_tgz_to_build.sh
@@ -52,7 +52,7 @@ temp=$(mktemp -d -t PyCascading-tmp-XXXXXX)
gzip -d <"$pycascading_dir/build/pycascading.tgz" >"$temp/pycascading.tar"
for j in "$@"; do
gzip -d <"$j" >"$temp/archive.tar"
- tar -A -f "$temp/pycascading.tar" "$temp/archive.tar"
+ tar -u -f "$temp/pycascading.tar" "$temp/archive.tar"
done
gzip -c <"$temp/pycascading.tar" >"$pycascading_dir/build/pycascading.tgz"
rm -rf "$temp"
View
3 remote_deploy.sh
@@ -66,7 +66,8 @@ Options:
-f <file> Copy file to the server together with main_script, but
do not bundle it up for submission. This option may be
repeated several times for multiple files. File names
- cannot start with a dot.
+ cannot start with a dot. This is useful if we need some
+ files to set up the flow, but not on every mapper/reducer.
-s <server> The name of the remote server where Hadoop is installed,
and the PyCascading scripts should be deployed to.

0 comments on commit 6c5660b

Please sign in to comment.