pcp
[Top] [All Lists]

Re: automatic derived metrics slowing down remote pcp clients, esp. pmlo

To: Mark Goodwin <mgoodwin@xxxxxxxxxx>
Subject: Re: automatic derived metrics slowing down remote pcp clients, esp. pmlogconf
From: "Frank Ch. Eigler" <fche@xxxxxxxxxx>
Date: Sat, 14 May 2016 21:08:57 -0400
Cc: pcp developers <pcp@xxxxxxxxxxx>
Delivered-to: pcp@xxxxxxxxxxx
In-reply-to: <13cab833-217d-a771-8deb-60ee5c23347f@xxxxxxxxxx>
References: <20160514193945.GC1418@xxxxxxxxxx> <d2f3f657-fe0b-c967-b7c0-e4603c9f5f9c@xxxxxxxxxx> <20160514220852.GD1418@xxxxxxxxxx> <13cab833-217d-a771-8deb-60ee5c23347f@xxxxxxxxxx>
User-agent: Mutt/1.4.2.2i
Hi -

> patch looks good - HOWEVER, 

To some extent, that's a separate matter; none of our pmlogconf
fragments invite the derived metrics for logging.

> thinking about this a bit more, ideally if a pmlogger config (or
> pmlogconf config) asks for a derived metric to be logged, it should
> expand this and log the leaf operand metrics. [...]

IMHO logging a derived metric - if specifically requested by a
pmlogger.conf file, should be exactly what pmlogger stores.  There is
no guarantee that any other client at any other place/time would have
the same definition for the derived metric.  Thus there's no reason to
assume that the values could be recomputed identically, even if the
individual inputs were stored.

- FChE

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