Hi Mark,
----- Original Message -----
> >
> > Can you try things with the attached patch?
>
> didn't make any difference. My system now has 36000+ containers (I'm
> determined to see what happens when docker creates the 65537th
> container, thus running out class b/c addresses, so I'm leaving my
> script running).
Oh, I see, so its continually extending the set ... right, yes, that
wont help there, only for the case where there were many containers
at agent start time (so, I'll test more and push this patch through
later).
This doesn't sound like a particularly realistic workload to optimise
for TBH, but it definitely needs to not hang things up...
> So we need a second thread for the initial scan, during which the
> main thread just returns PM_ERR_PMDANOTREADY for all requests. In
> reality, this many containers is a bit silly, so the patch for this
> is on my back-burner - let's move on to more important container work!
Indeed, as discussed - something like pmdarpm perhaps. So, the earlier
patch effectively does that just for the initial pmcd/pmda connect, but
not for subsequent PDUs... that's far more invasive change, and I agree,
lets come back to that once more fundamental container work is in place.
cheers.
--
Nathan
|