[Beowulf] Re: Rackable / SGI

[Tim Cutts]

On 5 Apr 2009, at 4:00 pm, Jason Riedy wrote:

>> A similar situation exists in the node management space, where
>> existing solutions like CFengine were pretty much ignored by HPC
>> people.
>
> Ha!  Cfengine was pretty much ignored by *everyone*, including
> its author for quite some time.  Promising (pun intended) the
> next great advance and not passing current maintenance to others
> loses users quickly.

Count one cfengine user here.

> Also, cfengine is (or was, when last I used it) designed to be a
> pull-based system that polls a configuration server.

It still is.  Our machines run it once a day.

>  The design
> was more focused on asynchronous updates, and I think most HPC
> folks would prefer a push model that updates everyone "at once."

I suppose we here don't mind the asynchronous nature, since we're  
mostly running embarrassingly parallel single threaded jobs.

> Cfengine had a push system, but to me it didn't feel like a good
> fit with the rest.

It doesn't really have a push system.  All it has is the ability to  
trigger all the nodes to pull at once.  Not something I use anyway.

> I'm more shocked that no one has written up using cfengine for
> managing laptops.

We use it for maintaining almost all of our Linux systems (2089  
systems, according to last night's report I got from it).  Not  
laptops, but certainly Linux desktops.

I don't think cfengine is perfect, by any means, but it does what we  
need for now.  cfengine3 is going to be such a big change from  
cfengine2 that I'm probably going to revisit the whole thing and see  
if I want to change to something else.

Tim
  


-- 
 The Wellcome Trust Sanger Institute is operated by Genome Research 
 Limited, a charity registered in England with number 1021457 and a 
 company registered in England with number 2742969, whose registered 
 office is 215 Euston Road, London, NW1 2BE.