One of our sysadmins uses a parallel interactive ssh utility. It opens separate windows for output and accepts input from one of the windows. When editing an config file where it would be a pain to build a sed script, this is better than pssh or equivalents. If the starting points are not identical this leads to *really* hosed results, of course. This problem afflicts anything of the sort and makes debugging problems really hard because the faults are so strange.
My own preference is to edit the files in question locally and then copy them to all targets. I have to understand the problem slightly better for that to work, but that seems to me to be a feature. On 4/29/08 2:52 PM, "Khalil Honsali" <[EMAIL PROTECTED]> wrote: > I was wondering if any of C3 or Capify offer the capability of doing > interactive distributed shell (if that ever makes sense), I am thinking of > the example of Yum's update on fedora, say without the default yes option. > > K. Honsali > > 2008/4/30 Bryan Duxbury <[EMAIL PROTECTED]>: > >> For commands on multiple machines, you can use Capistrano's shell utility. >> An added bonus is that you can write all sorts of more complicated processes >> using Ruby if you want to. >> >> www.capify.org >> >> -Bryan >> >> >> On Apr 29, 2008, at 1:36 PM, Bradford Stephens wrote: >> >> Greetings, >>> >>> I'm compiling a list of (free/OSS) tools commonly used to administer >>> Linux >>> clusters to help my company transition away from Win solutions. >>> >>> I use Ganglia for monitoring the general stats of the machines (Although >>> I >>> didn't get the hadoop metrics to work). I also use ntop to check out >>> network >>> performance (especially with Nutch). >>> >>> What do you all use to run your Hadoop clusters? I haven't found a good >>> tool >>> to let me run a command on multiple machines and examine the output, >>> yet. >>> >>> Cheers, >>> Bradford >>> >> >> > > > -
