Re: pcp updates: libpcp connection unsharing

----- Original Message -----
> > [...]
> > I'm all for code simplification, but the rationale here is not
> > compelling (don't get me wrong - I wrote the comment in context.c ...)
> > 
> >          * It is the topic of some debate as to whether PMCD connection
> >          * sharing is of much value at all, especially considering the
> >          * number of subtle and nasty bugs it has caused over time.  Do
> >          * not rely on this behaviour, it may well be removed someday.
> 
> > We need to properly understand the root cause of the problem;
> 
> Absolutely.  And until this problem is diagnosed (who's going to work
> on it?),

Not you?  I'll look at it eventually if nobody else does - Ken and Mark
have also fixed bugs in this area, anyone who puts their mind to it can
help though.  But we wont be making a change to 20+ year old code, on the
critical path in libpcp, with no evidence of improvement or understanding
of why it helps (if it even helps).

Have you written any test code so far to isolate and attempt to reproduce
the problem outside of pmwebd?  (inducing timeouts, interleaving destroy-
context with other calls, etc)?  Could you share it please, if so, so that
other folk can help out and try to understand the root cause too.  If not,
doing that would be quite helpful & would form the basis of those QA tests
Mark mentioned.

> > If it helps any, there is an undocumented hostspec flag
> > ("exclusive") which can be used here [...]
> 
> It would need to be added to all the multiple-context pmapi clients,
> including pmmgr, pmwebd, pmchart, pmie, and probably others.

By "used here" I meant here in triaging/testing/debugging, trying to get to
the bottom of the problem.


So its a NAK to this change from me until the problem is understood; let's
diagnose root causes and make informed decisions.

As to the valgrind threading reports, same thing - we need to see specific
issues being resolved there and explain how, why, etc.  They could well be
simple things that are easily fixed (if PMAPI threaded tools eventuate) or
non-issues in practice - who knows?  Not us, not yet.

cheers.

--
Nathan

-=-=-=-=-=-=-=-=-=-=-=-
pcp mailing list
[email protected]
https://groups.io/g/pcp/messages
-=-=-
Groups.io Links:

You receive all messages sent to this group.

View/Reply Online (#14666): https://groups.io/g/pcp/message/14666
View All Messages In Topic (10): https://groups.io/g/pcp/topic/3038481
Mute This Topic: https://groups.io/mt/3038481?uid=174580
New Topic: https://groups.io/g/pcp/post

Change Your Subscription: https://groups.io/g/pcp/editsub?uid=174580

Group Home: https://groups.io/g/pcp
Contact Group Owner: [email protected]

Terms of Service: https://groups.io/static/tos

Unsubscribe: https://groups.io/g/pcp/leave/354243/563757577/xyzzy
-=-=-=-=-=-=-=-=-=-=-=-