Hi Guille, Thanks for catching this. You are right, the slaves.sh should strip out the prefix and port for collector file to be consistent between documentation and the code. JIRA-455 filed. :)
Regards, Eric On 2/19/10 5:51 AM, "Guillermo Pérez" <bi...@tuenti.com> wrote: > The documentation explain to put in the conf/collectors file: > > http://<collector1HostName>:<collector1Port>/ > http://<collector2HostName>:<collector2Port>/ > http://<collector3HostName>:<collector3Port>/ > > But that file is also used by the bin/start-all.sh script, that simply > tries to ssh to that. Of course, ssh to http://something... doesn't > work. > > I have modified the slaves.sh file for striping http:// and :<port> > from that file for getting the host to ssh to. I hope this may help > other people (maybe I'm doing something wrong?) > > for slave in `cat "$HOSTLIST" | sed -e 's/^http:\/\///;s/:[0-9]*.*//'`; do
