Always-on recursive application of all rebar commands causes too many
issues. Recursive application is required for:
1. dealing with dependencies: get-deps, update-deps, and compile of deps
right after get-deps or update-deps
2. projects with a riak-like apps/ project structure and dev process
The vast majority of projects are not structured like riak. Therefore,
moving forward it's best to (by default) restrict recursive behavior to
dealing with deps. This commit does that and also adds command line and
rebar.config options for controlling or configuring recursion. Also, we
introduce two meta commands: prepare-deps (equivalent to rebar -r
get-deps compile) and refresh-deps (equivalent to rebar -r update-deps
compile). riak-like projects can extend the list of recursive commands
(to include 'eunit' and 'compile') by adding
{recursive_cmds, [eunit, compile]} to rebar.config.
When a node is configured with -sname in app.config or sys.config the
REMSH_NAME_ARG and REMSH_REMSH_ARG arguments are incorrect due the
assumption that the node name always contains '@'.
To fix the bug, the script tries to find '@' and then compose the
arguments acording to node name type. If in long name mode the
script can't compose a correct node name exits with warning.
It is a portable version of the realpath(1) utility that you can find on
Mac OS X and FreeBSD (see also The Open Group Base Specifications Issue
6, IEEE Std 1003.1).
Without the -P flag, pwd(1) might return different values when the
current path contains one or more symlinks, depending on how you got
into the current directory.
In simplenode.runner, this may cause PIPE_DIR to have different values
on each use, which will make it impossible to connect to the running
node unless you guess the correct path yourself.
When developing Riak, we have found bugs and other issues due
to the number of platforms we support.
Here is an overview of the changes:
- Fix command-line syntax for commands to work on *BSD / Sun
- Add chkconfig and getpid to nodetool
- Replace platform specific 'kill' commands with a nodetool
getpid method
- Fix RUNNER_USER settings to work on *BSD
This patch contains two additions to simplenode.runner:
1. Check if vm.args exists in CWD; if so, use it. This makes it
easier to start multiple concurrent nodes on a single machine
from one rebar-created release (starting each node from its
own directory, with its own copy of vm.args and e.g.
sys.config, log directory, database directory, etc.)
2. Add the targets start_boot <file> and console_boot <file>.
This is used to select a different boot script. The 'setup'
application (http://github.com/esl/setup) builds a special boot
script for installation (all apps loaded but not started,
making it possible to run install hooks with the full code path
in place).
* Typos and other minor bug fixes
* Add "attach" command (via -remsh)
* Add "upgrade" command to install a hot upgrade package
* Add "upgrade" subcommand to UNIX runner script as well
bin/<release_name> console -kernel test '[{"a","b"}]' fails, because
when $@ is inserted into CMD [{"a","b"}] is interpreted.
Fix the same issue for 'start' command and quote all arguments.
