[collectd] rrdcached Plugin

[Zhou, Yuantai]

Hi All,
We are testing the RRDCached plugin but are seeing a number of errors in syslog from collectd servers:
rrdcached[14880]: queue_thread_main: rrd_update_r (/var/lib/rrdcached/db/collectd/xxxxxxxxx/disk-dm-0/disk_time.rrd) failed with status -1. (/var/lib/rrdcached/db/collectd/xxxxxxxxxxx/disk-dm-0/disk_time.rrd: illegal attempt to update using time 1367363996 when last update time is 1367364056 (minimum one second step))

collectd[32487]: uc_update: Value too old: name = xxxxxxxxxxx/cpu-3/cpu-user; value time = 1367364236.000; last cache update = 1367364236.000;

rrdcached plugin: rrdc_update (/var/lib/rrdcached/db/collectd/xxxxxxxxxxx/df/df-root.rrd, [1367364236:98146201600.000000:1892331495424.000000], 1) failed with status -1.


The RRDCached plugin configurations is as follows:

<Plugin rrdcached>
        DaemonAddress "unix:/var/run/rrdcached.sock"
        DataDir "/var/lib/rrdcached/db/collectd"
        CollectStatistics true
</Plugin>

<Plugin rrdtool>
        DataDir "/var/lib/rrdcached/db/collectd"
        CacheTimeout 60
        CacheFlush 3600
        WritesPerSecond 1800
        RandomTimeout 30
</Plugin>

We do see that the RRD files (at least some or most of them) are being updated (with new timestamps).

The rrdcached process seems to have 6 threads:
25379 root      20   0  148m 1236  464 S    0  0.0   0:00.11 rrdcached
25380 root      20   0  148m 1236  464 S    0  0.0   0:00.00 rrdcached
25381 root      20   0  148m 1236  464 S    0  0.0   0:00.00 rrdcached
25382 root      20   0  148m 1236  464 S    0  0.0   0:00.00 rrdcached
25383 root      20   0  148m 1236  464 S    0  0.0   0:00.00 rrdcached
25384 root      20   0  148m 1236  464 S    0  0.0   0:00.00 rrdcached

Any suggestions on what we did wrong?

Thanks.
Yuantai

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.verplant.org/pipermail/collectd/attachments/20130430/99662b4b/attachment.html>