pcp
[Top] [All Lists]

Release status (was Re: [pcp] pmRegisterDerived return values)

To: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Subject: Release status (was Re: [pcp] pmRegisterDerived return values)
From: Nathan Scott <nathans@xxxxxxxxxx>
Date: Wed, 27 Jan 2016 00:02:26 -0500 (EST)
Cc: pcp developers <pcp@xxxxxxxxxxx>
Delivered-to: pcp@xxxxxxxxxxx
In-reply-to: <56A6FA88.1020705@xxxxxxxxxxxxxxxx>
References: <569CB025.4070603@xxxxxxxxxx> <569CB7C5.7030803@xxxxxxxxxxxxxxxx> <56A52ED5.6040603@xxxxxxxxxxxxxxxx> <1086103317.13561518.1453675424866.JavaMail.zimbra@xxxxxxxxxx> <56A6FA88.1020705@xxxxxxxxxxxxxxxx>
Reply-to: Nathan Scott <nathans@xxxxxxxxxx>
Thread-index: yA4gm4DiujrkcwGNO6LZ5+c753G5qQ==
Thread-topic: Release status (was Re: [pcp] pmRegisterDerived return values)

----- Original Message -----
> > [...]
> > Hmmm, I suspect we should do both, based on the way all the other libpcp
> > parser interfaces do error handling.
> 
> OK, I agree.  I've done the parser work (commit coming) to ensure when
> pmRegisterDerived() (and friends) return with an error, there is
> something useful in the error message buffer in all cases.
> 

Got it, thanks Ken.  I've merged that and a handful of other small items;
I'm expecting that should be it for this release.

I've enabled PMCD_ROOT_AGENT in /etc/sysconfig/pmcd in this final set too,
as its been QA'ing nicely for Lukas and I since it went in (and getting a
pmcd SIGHUP able to restart PMDAs without restarting pmcd once again is
an important fix for some folk).

There is a slight flappiness to qa/110 which is definitely timing related
(the af_unix install there, done twice, sometimes fails once) - you might
well see this one too, I'll keep working on making it deterministic.  But
other than that, QA is looking good here & on the buildbots.

cheers.

--
Nathan

<Prev in Thread] Current Thread [Next in Thread>