pcp
[Top] [All Lists]

Re: [pcp] FC19 et al QA failures ...

To: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Subject: Re: [pcp] FC19 et al QA failures ...
From: Nathan Scott <nathans@xxxxxxxxxx>
Date: Mon, 2 Mar 2015 16:55:34 -0500 (EST)
Cc: PCP <pcp@xxxxxxxxxxx>
Delivered-to: pcp@xxxxxxxxxxx
In-reply-to: <54F4CBCE.2040406@xxxxxxxxxxxxxxxx>
References: <54F4CBCE.2040406@xxxxxxxxxxxxxxxx>
Reply-to: Nathan Scott <nathans@xxxxxxxxxx>
Thread-index: GS5IAsjH4X7p/sDSb8zyp9c+IbC9AA==
Thread-topic: FC19 et al QA failures ...
Hi Ken,

----- Original Message -----
> [...]
> What is the recommended way to address this?
> 

(I'm using sendmail here, FWLIW)

> Note that installing postfix (with no configuration) has no impact ... the
> original problem persists.
> 
> On the other hand, removing postfix and installing sendmail resolves the
> issue.

I wonder if we should rethink our strategy of sending mail when unresolvable
problems occur?  An alternative might be to create a logfile with (sometimes
large) body of the mail in a relevant location, then log a one-line summary
to syslog?  (and do away with all reliance on sending mail)?

In recent production environments I've worked, this'd have been a preferable
strategy in general, as it would've fitted in more neatly with the central
logging and alerting systems in use.  But, YMMV, maybe other folks prefer to
get mail for this kind of thing still... not sure.

cheers.

--
Nathan

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