I guess pmlogconf escaped at some point after the initial open source
PCP release ... it certainly is in something of a "state".
I'd be willing (happy is probably a bit over the top) to take on some
remediation in this area.
Moving from hard-coded to the contents of the file store (a la pmieconf)
for the default groups and group->metrics list mappings is probably not
too difficult.
The largest issue I see here is that the host where pmlogconf is run may
or may not be the host where the pmcd that pmlogger will connect to is
to be run, and even worse the metrics available on the local host and
the host where pmlogger gets its metrics from may be very different.
So it is not even clear what metrics should be in the external
configuration file(s).
We almost need a command line option to specify something about the
platform where pmcd will be running, like
- a generic platform, e.g. Irix, Linux, Windows, Solaris, MacOSX and/or
- -h hostname (and then use a Larry Wall style algorithm to deduce that
hostname smells like, er, Linux)
Now one could see how this might work for an initial invocation to set
up the skeleton. What is less clear is what happens subsequently, e.g.
imagine foo.bar.mumble is in the control file, but is not present on the
pmcd host when pmlogconf is run first. Later on, after an O/S upgrade
or PMDA install, foo.bar.mumble is available on the pmcd host. What is
the desired behaviour?
It may take a bit of discussion before we finalize the specs for what
pmlogconf should really do.
On Thu, 2010-06-10 at 15:28 +1000, Nathan Scott wrote:
> Hi Ken,
>
> I was taking someone through whats involved in setting up a pmlogger
> farm the other day, and in the process mentioned pmlogconf (which I'd
> not really used much before). This was in the context of "how would
> one setup a logging farm for a many-node hadoop cluster".
>
> Its in a bit of a state, not sure if you want to take on renovations
> in that area. :) I know Mark was looking for something similar for
> getting loggers running in Fedora/RHEL out-of-the-box. The Aconex
> production environment uses a comparable process for setting up our
> loggers, we might be convinced to convert if pmlogconf got a few more
> features (in hindsight, we should've taken this on when we set that
> up initially ... oh well).
>
> Here's a few things that come to mind in terms of work-needed:
> - its very IRIX-centric currently, will need to cull old metrics and
> introduce more appropriate ones from other platforms
> - the metrics are all hard-coded in the script, a plug-in model more
> along the lines of pmieconf would seem better
> - ability to probe a host (local/remote) for available metrics to
> automatically setup logging groups
>
> If you (or anyone) take this on, happy to help with specific metrics
> and processes we use to configure our production loggers for Linux &
> Windows servers.
>
> cheers.
>
|