pcp
[Top] [All Lists]

Re: [pcp] QA regressions with cgroup.groups

To: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Subject: Re: [pcp] QA regressions with cgroup.groups
From: Nathan Scott <nathans@xxxxxxxxxx>
Date: Wed, 6 Aug 2014 19:37:19 -0400 (EDT)
Cc: PCP <pcp@xxxxxxxxxxx>
Delivered-to: pcp@xxxxxxxxxxx
In-reply-to: <53E2ADCD.6050704@xxxxxxxxxxxxxxxx>
References: <53E2ADCD.6050704@xxxxxxxxxxxxxxxx>
Reply-to: Nathan Scott <nathans@xxxxxxxxxx>
Thread-index: yiDFCmHrGC1ZzgnzqaCJ4kEANbfyDQ==
Thread-topic: QA regressions with cgroup.groups
Hi Ken,

----- Original Message -----
> I'm back in the QA pool.
> 
> I am seeing a number of (apparently QA) regressions where this is the
> root cause
> 
> cgroup.groups: pmLookupName: Unknown metric name
> 
> Has something changed here in the recent past?

Yes, there's been fairly extensive changes in cgroups recently.  I'm
not seeing these tests fail though - probably cos I'm not running on
machines without cgroup support.

> Tests failing for me on Debian caused by this: 183, 184, 185 and 361
> 
> Is there some easy setup to make cgroup.groups.* known in the PMNS via
> the proc PMDA?
> 

This subtree is populated when cgroup subsystems exist - I guess these
are older kernels with no cgroups?  (hmm I wonder if this is the right
error code for this situation?).

One part of the last round of changes was to make cgroup.groups. the
dynamic non-leaf PMNS entry, whereas previously it was cgroup. - and
that was *always* populated with the cgroup.mounts and cgroup.subsys
(fixed) hierarchies.  Could be some latent issue with dynamic names?
Or maybe I've just (ab)used them in a new way.

I notice mmv always does this "pre-populate with a few fixed names"
for its dynamic namespace use, and the Linux kernel interrupt names
(also dynamic) will always have a starting set... hmmm.

cheers.

--
Nathan

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