[collectd] errors with Filter subsystem: Built-in target `write'

[Jesse Reynolds]

:-) 

Thanks Andrew, 

Yeah, I was aware we could do this. If the warnings really are benign, then this is OK. 


On 18/03/2014, at 14:34:48, "GRAY Andrew G (SPARQ)" <andrew.gray1 at sparq.com.au> wrote:

> Jesse,
> Here’s a working fix for you, kludgy I know …. Doesn’t answer why, but does get you a working system.
>  
> /etc/collectd.conf
> Change syslog level from info to warning.
>  
> <Plugin syslog>
>                 LogLevel warning
> </Plugin>
>  
>  
> Regards,
>  
> Andrew Gray.
> RHCSA, Professional Unix Administration.
>  
> From: collectd-bounces at verplant.org [mailto:collectd-bounces at verplant.org] On Behalf Of GRAY Andrew G (SPARQ)
> Sent: Tuesday, 18 March 2014 1:03 PM
> To: 'Jesse Reynolds'
> Cc: collectd at verplant.org
> Subject: Re: [collectd] errors with Filter subsystem: Built-in target `write'
>  
> Hi Jesse,
>  
> OK, shiny new 1.48 installed … no change.
> What I have tried.
>  
> # Explicitly set write target:
> <Chain “PostCache”>
>                 <Target “write”>
>                                 Plugin “rrdcached”
>                 </Target>
> </Chain>
>  
> No change … apart from “Dispatching value to all write plugins” changing to “Dispatching value to ‘rrdcached’ plugin failed with status -1”
>  
> Notes that the rrdc_update always fails the same test for every machine. Currently that would be tcp_connections, but before I commented it out,
> It was the df (disk free) test that was failing on every server.
>  
> It does not seem to stop the graphs being created, just fills the error log.
>  
>  
> Regards,
>  
> Andrew Gray.
> RHCSA, Professional Unix Administration.
>  
> From: Jesse Reynolds [mailto:jesse at bulletproof.net] 
> Sent: Tuesday, 18 March 2014 10:36 AM
> To: GRAY Andrew G (SPARQ)
> Cc: david van zeebroeck; collectd at verplant.org
> Subject: Re: [collectd] errors with Filter subsystem: Built-in target `write'
>  
> Interesting, 
>  
> I am running rrdtool 1.4.8 (compiled afresh) and collectd 5.4.1 (compiled afresh also) on Ubuntu Precise 12.04. 
>  
> Perhaps it's to do with the version of rrdtool. Which version of rrdtool is your rrdcached using, Andrew? 
>  
>  
> On 18/03/2014, at 09:57:27, "GRAY Andrew G (SPARQ)" <andrew.gray1 at sparq.com.au> wrote:
>  
> 
> I’m seeing this too,  I have tried mixing and matching different versions. 5.0.3, 5.1.0, 5.4.1. Recompiled my own collectd rpms. No change.
> I’m down to trying another version of rrdtools (the one I’m using is off rpmfind.net, Rrdtools 1.4.7-1.el6.rfx.x86_64.rpm)
>  
> I only seem to get this when I run rrdcached, when running rrdtools plugin the logs are quiet.
>  
> Regards,
>  
> Andrew Gray.
> RHCSA, Professional Unix Administration.
>  
> From: collectd-bounces at verplant.org [mailto:collectd-bounces at verplant.org] On Behalf Of Jesse Reynolds
> Sent: Monday, 17 March 2014 9:11 PM
> To: david van zeebroeck
> Cc: collectd at verplant.org
> Subject: Re: [collectd] errors with Filter subsystem: Built-in target `write'
>  
> Did you work this out David? 
>  
> I'm seeing the same thing on a test collectd 5.4.1 server I've just set up. 
>  
> The server is busy (>500 packets per second arriving on it) and there are heaps of these errors being logged. 
>  
> Anyone got any ideas? 
>  
> Eg:
>  
> Mar 17 22:01:31collectd-03 collectd[16278]: uc_update: Value too old: name = foo.example/snmp/if_packets-FastEthernet0_22; value time = 1395054090.000; last cache update = 1395054090.000;
> Mar 17 22:01:31collectd-03 collectd[16278]: Filter subsystem: Built-in target `write': Dispatching value to all write plugins failed with status -1.
>  
> I am using the v5 upgrade target. All the clients are version 4 of collectd, or SSC Serv. 
>  
>  
> On 15/10/2013, at 20:16:20, david van zeebroeck <david at analytics.brusselsairport.be> wrote:
> 
> 
> 
> hello
> i've installed collectd 5.4
> everything run fine except i get a lot of error like these from some clients 
>  
> [2013-10-15 11:43:45] Filter subsystem: Built-in target `write': Dispatching value to all write plugins failed with status -1.
> [2013-10-15 11:43:45] uc_update: Value too old: name = elara/disk-dm-0/disk_merged; value time = 1381830225.000; last cache update = 1381830225.000;
> [2013-10-15 11:43:45] Filter subsystem: Built-in target `write': Dispatching value to all write plugins failed with status -1.
> [2013-10-15 11:43:45] uc_update: Value too old: name = elara/disk-dm-1/disk_octets; value time = 1381830225.000; last cache update = 1381830225.000;
>  
> i've checked that collectd is not running twice
> also ntp is running on both server & client and in sync
>  
> i'm not using rrdcached
>  
> my rrdtool config looks like this :
>  
> LoadPlugin rrdtool
> <Plugin rrdtool>
>        DataDir "/var/lib/collectd/rrd/"
>        CreateFilesAsync false
>        CacheTimeout 120
>        CacheFlush   900
>        WritesPerSecond 50
> </Plugin>
> <target "write">
> plugin "rrdtool"
> </target>
>  
> any clues anyone?
> _______________________________________________
> collectd mailing list
> collectd at verplant.org
> http://mailman.verplant.org/listinfo/collectd



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.verplant.org/pipermail/collectd/attachments/20140318/52d746c2/attachment-0001.html>