pcp
[Top] [All Lists]

Re: [pcp] qa/1111 failing pretty much every place it is run

To: Alec Ten Harmsel <alec@xxxxxxxxxxxxxxxxxx>, Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Subject: Re: [pcp] qa/1111 failing pretty much every place it is run
From: Nathan Scott <nathans@xxxxxxxxxx>
Date: Sun, 17 Jul 2016 23:44:31 -0400 (EDT)
Cc: PCP <pcp@xxxxxxxxxxx>
Delivered-to: pcp@xxxxxxxxxxx
In-reply-to: <20160716204609.GB810@apio>
References: <57858370.8030700@xxxxxxxxxxxxxxxx> <131083523.5554077.1468368582128.JavaMail.zimbra@xxxxxxxxxx> <747918044.5554823.1468369188587.JavaMail.zimbra@xxxxxxxxxx> <20160716204609.GB810@apio>
Reply-to: Nathan Scott <nathans@xxxxxxxxxx>
Thread-index: EGJNbqgqvNP31BpHp3jk9nV7rKcNUw==
Thread-topic: qa/1111 failing pretty much every place it is run

----- Original Message -----
> On Tue, Jul 12, 2016 at 08:19:48PM -0400, Nathan Scott wrote:
> > 
> > We might want to have some cleaner error reporting in pcp2influxdb there -
> > I guess it is feasible influxdb could restart and cause the same trace?  If
> > so a less alarming end-user diagnostic would be preferable.
> 
> I've attached a patch that adds some more helpful error messages. It now
> prints warnings in the following situations that it did not before:
> 
> * The python requests package is not installed (also hard exits after
>       printing the message)
> * The HTTP request returns 404 (most likely the requested database does
>       not exist)
> * The connection fails
> 

Looks good to me, thanks for following up Alec!  I'll merge shortly, and
I'm expecting you'll already be seeing zero 1111 failures now, Ken, from
earlier updates - let us know if not?  Taa.

cheers.

--
Nathan

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