OK, this looks like getting pretty ugly.
Consider the following 4 outputs from pminfo ...
kenj@bozo:~/src/pcp$ pminfo -n /var/lib/pcp/pmns/root -L -m mmv
mmv PMID: 70.*.*
kenj@bozo:~/src/pcp$ pminfo -n /var/lib/pcp/pmns/root -m mmv
mmv PMID: 70.*.*
kenj@bozo:~/src/pcp$ pminfo -L -m mmv
mmv PMID: 70.*.*
kenj@bozo:~/src/pcp$ pminfo -m mmv
mmv.test.strings PMID: 70.1.6
mmv.test.string PMID: 70.1.5
mmv.test.interval PMID: 70.1.4
mmv.test.indom PMID: 70.1.3
mmv.test.discrete PMID: 70.1.2
mmv.test.counter PMID: 70.1.1
mmv.debug PMID: 70.0.1
mmv.reload PMID: 70.0.0
According to my original scope, these are all correct.
Nathan would like the third one to produce the same output as the last
one.
If this was to happen, what is the "correct" output for the first and
second cases?
I'd like to see some justification, not just votes please.
Of course pminfo is a worst case scenario because it uses unusual logic
to try very hard to resolve PMNS-only requests without reference to pmcd
(or pmdas).
On Mon, 2010-04-12 at 13:52 +1000, Ken McDonell wrote:
> This is pretty much outside the scope of what I was aiming for with the
> dynamic metrics, although it is not unreasonable.
>
> It will mean even more code being copied and changed from pmcd to libpcp
> as the current code in the library has no support for making the needed
> DSO calls.
>
> So it is not a simple change.
>
> The mmv pmda will sneak in on the back of the default dso table for
> PM_CONTEXT_LOCAL, but how will you get the aconex pmda into the dso
> table (pminfo -L -K 123,aconex/pmda_aconex.so,aconex_init ... in the
> bash stuff)?
>
> On Mon, 2010-04-12 at 11:56 +1000, nathans@xxxxxxxxxx wrote:
> > Hi Ken,
> >
> > We're a long way down the path of converting all of our production
> > environment over to using MMV and the dynamic namespace extensions.
> > All going swimmingly so far. Just this morning we came across an
> > issue with the bash metric name auto-completion. This uses pminfo
> > in local context mode (-L) to provide a list of metric names which
> > the shell can use for name completion with several pcp commands.
> >
> > In our namespace we have MMV:*:* (a top level mmv namespace entry)
> > and ACONEX:*:* as well (likewise, at the top level). In local mode
> > pminfo does not traverse, just reports "mmv" and "aconex".
> >
> > We can change the bash code to not use -L, but that seems like its
> > just avoiding the real problem - should local context mode be able
> > to descend into these parts of the namespace? I woulda thought it
> > should, just like regular pmcd mode...?
> >
> > cheers.
> >
>
>
> _______________________________________________
> pcp mailing list
> pcp@xxxxxxxxxxx
> http://oss.sgi.com/mailman/listinfo/pcp
|