pcp
[Top] [All Lists]

Re: [pcp] Using PCP and Zabbix

To: Trey Dockendorf <treydock@xxxxxxxx>, Marko Myllynen <myllynen@xxxxxxxxxx>
Subject: Re: [pcp] Using PCP and Zabbix
From: Nathan Scott <nathans@xxxxxxxxxx>
Date: Tue, 26 Jan 2016 19:31:07 -0500 (EST)
Cc: pcp@xxxxxxxxxxx
Delivered-to: pcp@xxxxxxxxxxx
In-reply-to: <CAJ=TzikkukNBRTZggzuu20cK-wXAfY5-vAPHgVbjsywpkM0vBQ@xxxxxxxxxxxxxx>
References: <CAJ=TzikkukNBRTZggzuu20cK-wXAfY5-vAPHgVbjsywpkM0vBQ@xxxxxxxxxxxxxx>
Reply-to: Nathan Scott <nathans@xxxxxxxxxx>
Thread-index: B7sN0z+JRIpIqdzhFe/THE47SSlSEw==
Thread-topic: Using PCP and Zabbix
Hi Trey,

----- Original Message -----
> I'm curious how others are managing PCP and Zabbix. The PCP docs for zbxpcp
> mention that pmcd needs to be running before the zabbix-agent. The issue
> that I'm curious about is how the order is achieved in CentOS 6 and system
> V. Installing PCP from bintray and Zabbix from EPEL results in pmcd (95)
> starting in runlevel 3 after zabbix-agent (86).
> 
> The only maintainable solution I can think of is to put something like
> 'service zabbix-agent restart' in /etc/rc.local.
> 
> I'm curious if maybe there's some more elegant solution I'm just overlooking.
> 

None that I know of so far, but Marko is the zbxpcp guru, perhaps he has one.
The zbxpcp code could possibly acquire ability to delay pmNewContext(3) and/or
use the pmReconnectContext(3) libpcp API to tackle this.

cheers.

--
Nathan

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