[erlang-questions] Possible bug in slave:start?

Eider Oliveira eider.oliveira@REDACTED
Fri Feb 22 22:09:42 CET 2008


I have a piece of code which runs differently if the remote node is started
through slave:start:
eider@REDACTED:~/bench> erl +K true -rsh ssh -sname bench -setcookie
bench_cookie
Erlang (BEAM) emulator version 5.6 [source] [smp:2] [async-threads:0] [hipe]
[kernel-poll:true]

Eshell V5.6  (abort with ^G)
(bench@REDACTED)1> slave:start('igaessi14-b',bench1,lists:concat(["
-setcookie ",erlang:get_cookie()])).
{ok,'bench1@REDACTED'}
(bench@REDACTED)2> mnesia:create_schema([node()|nodes()]).
{error,{"Cannot install fallback",

 {file_error,"/home/eider/bench/Mnesia.bench1@REDACTED/FALLBACK.TMP.BUPTMP",
                    enoent}}}
(bench@REDACTED)3> q().
ok


The problem is the /home/eider/bench/Mnesia.bench1@REDACTED directory was
created on the first machine (igaessi12-b):

eider@REDACTED:~/bench> ls -l Mne*
drwxr-xr-x  2 eider implantacao     48 2008-02-22 17:30
Mnesia.bench1@REDACTED
drwxr-xr-x  2 eider implantacao    112 2008-02-22 17:30
Mnesia.bench@REDACTED

If erlang is started at the second machine (igaessi14-b) when the previous
code run, it works correctly:

eider@REDACTED:~/bench> erl +K true -rsh ssh -sname bench -setcookie
bench_cookie
Erlang (BEAM) emulator version 5.6 [source] [smp:2] [async-threads:0] [hipe]
[kernel-poll:true]

Eshell V5.6  (abort with ^G)
(bench@REDACTED)1> slave:start('igaessi14-b',bench1,lists:concat(["
-setcookie ",erlang:get_cookie()])).
{error,{already_running,'bench1@REDACTED'}}
(bench@REDACTED)2> mnesia:create_schema([node()|nodes()]).
ok
(bench@REDACTED)3> q().
ok

-- 
Eider Oliveira

Site: http://eider.eti.br
Blog: http://eider.eti.br/Home/Blog/Blog.html

"If I had more time, I'd have written you a shorter letter." Pascal
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://erlang.org/pipermail/erlang-questions/attachments/20080222/bd4f5ffc/attachment.htm>


More information about the erlang-questions mailing list