Re: [Ganglia-developers] bootstrapping for 3.1.X series and 3.2.X

2010-01-08 Thread Kostas Georgiou 
On Thu, Jan 07, 2010 at 02:36:38PM +, Daniel Pocock wrote:

  then you are going to need either 2 public resources for all release 
  managers
  to use consistently or a coordinate release process were the package is
  generated and then independently binary packages are added to it before
  the announcement (which also means we have to agree on what is going to be
  used for building those RPM packages).
 

 Not necessarily - it may be sufficient to provide some scripts that 
 release managers can use, as long as the hostnames can be easily 
 configured somewhere.  Then the release manager just needs to have the 
 necessary machines available, but that is not so difficult either thanks 
 to Xen, VMware, etc.

With mock[1] you probably don't need anything beyond one machine, mock
basically runs rpmbuild inside a minimal chroot (which also installs for
you) of a Fedora/Centos (might also be usable for Suse) release that you
want and the rpm dependencies of the package. Given that most developers
will have a Fedora/Centos/RHEL/Debian machine building rpms shouldn't be
a problem really.

Cheers,
Kostas

[1] http://fedoraproject.org/wiki/Projects/Mock

--
This SF.Net email is sponsored by the Verizon Developer Community
Take advantage of Verizon's best-in-class app development support
A streamlined, 14 day to market process makes app distribution fast and easy
Join now and get one step closer to millions of Verizon customers
http://p.sf.net/sfu/verizon-dev2dev 
___
Ganglia-developers mailing list
Ganglia-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ganglia-developers

[Ganglia-developers] backporting r2133 - load_metric_modules()

2010-01-08 Thread Daniel Pocock 

I've just looked at r2133.

I notice that there is a call to load_metric_modules() within the first 
conditional block (args_info.metrics_flag).

Therefore, to solve the problem addressed by r2133, I believe one of the 
following alternatives is needed:

a) put a call to load_metric_modules() within the conditional block for 
bandwidth_flag

b) call load_metric_modules() before both conditional blocks (this is 
what I prefer)

Any comments/objections?

  if(args_info.metrics_flag)
{
  load_metric_modules();
  initialize_scoreboard();
  setup_metric_callbacks();
  print_metric_list();
  fflush( stdout );
  exit(0);
}

  load_metric_modules();

  if(args_info.bandwidth_flag)
{
  double bytes_per_sec;
  setup_metric_callbacks();
  bytes_per_sec = setup_collection_groups();
  fprintf(stdout, %f bytes/sec\n, bytes_per_sec);
  exit(0);
}



--
This SF.Net email is sponsored by the Verizon Developer Community
Take advantage of Verizon's best-in-class app development support
A streamlined, 14 day to market process makes app distribution fast and easy
Join now and get one step closer to millions of Verizon customers
http://p.sf.net/sfu/verizon-dev2dev 
___
Ganglia-developers mailing list
Ganglia-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ganglia-developers

[Ganglia-developers] backports for 3.1.6 - testing

2010-01-08 Thread Daniel Pocock 


I've just done some backports for 3.1.6 (see log below), there may be a 
couple of others to come too

- are there any more backports that people would like?

- it's probably time for people to start testing the monitor-core-3.1 
branch with the view that we will tag 3.1.6 in the next 1-2 weeks

Some other things that have been discussed for 3.1.6 on the list:

- bootstrapping - we are deciding whether to use Debian 5 or the same 
environment as 3.1.2

- dropping the release names, and just using a release number

- git - maybe this won't happen until later


r2190 | d_pocock | 2010-01-08 15:47:40 + (Fri, 08 Jan 2010) | 1 line

Rewrite gmetad sleep code in various places to use apr, remove magic 
numbers, sleep as a percentage of the step rather than an absolute 
random adjustment (backport r2106, r2107, r2108)

r2191 | d_pocock | 2010-01-08 15:49:04 + (Fri, 08 Jan 2010) | 1 line

gmetad: workaround broken apr-1 headers depending incorrectly on off64_t 
(backport r2188)

r2192 | d_pocock | 2010-01-08 15:51:53 + (Fri, 08 Jan 2010) | 1 line

libmetrics: freebsd fixes for crashes and disk statistics (BUG153) 
(backport r2126)

r2193 | d_pocock | 2010-01-08 16:55:41 + (Fri, 08 Jan 2010) | 1 line

documentation: contrib description for ganglia-hosts-lowercase (backport 
r2134)

r2194 | d_pocock | 2010-01-08 16:58:25 + (Fri, 08 Jan 2010) | 1 line

web: update documentation for case_sensitive_hostnames to reflect it is 
a bool (backport r2135)

r2195 | d_pocock | 2010-01-08 17:01:07 + (Fri, 08 Jan 2010) | 1 line

build: remove hack to force C99 support with Sun Studio (backport r2148)

r2196 | d_pocock | 2010-01-08 17:03:09 + (Fri, 08 Jan 2010) | 1 line

Annotate that send_metadata_interval is in seconds (backport r2151)

r2197 | d_pocock | 2010-01-08 17:04:23 + (Fri, 08 Jan 2010) | 1 line

libmetrics: solaris: avoid conflict with ncpus extern in old versions of 
sys/cpuvar.h (backport r2153)

r2198 | d_pocock | 2010-01-08 17:06:14 + (Fri, 08 Jan 2010) | 1 line

build: check for Python.h explicitly (backport r2154)

r2199 | d_pocock | 2010-01-08 17:09:22 + (Fri, 08 Jan 2010) | 1 line

documentation: remove references to GNU buildchain as a hard dependency 
(backport r2155)

r2201 | d_pocock | 2010-01-08 17:17:43 + (Fri, 08 Jan 2010) | 1 line

gmetad: fortify write_data_to_rrd against possible overflows (backport 
r2159 and r2200)

r2202 | d_pocock | 2010-01-08 17:22:31 + (Fri, 08 Jan 2010) | 1 line

web: use tables instead of forced breaks for aligning metrics in host 
view (backport r2161)

r2203 | d_pocock | 2010-01-08 17:25:32 + (Fri, 08 Jan 2010) | 1 line

web: only add new row if there are more metrics in section (backport r2162)



--
This SF.Net email is sponsored by the Verizon Developer Community
Take advantage of Verizon's best-in-class app development support
A streamlined, 14 day to market process makes app distribution fast and easy
Join now and get one step closer to millions of Verizon customers
http://p.sf.net/sfu/verizon-dev2dev 
___
Ganglia-developers mailing list
Ganglia-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ganglia-developers