If you don’t miss any historical data in your graphs, you can get rid of this message by stopping your site for a brief moment, running cmk -v --convert-rrds --delete-rrds and then starting the site again.
The command will convert any historical perfdata remaining in the old format and convert it into the new format.
The possible problem is that you have a system with “mixed” RRDs at the moment.
Can you have a look at your “~/var/pnp4nagios/perfdata/” folder and check how many data is there left after the convert?
If you check this for one if the entries from the log you should find an RRD file but no corresponding XML i think.
I also experienced the problem described above on an upgraded 2.0.0p24 (CEE).
I did try
Afterwards I still have a folder per monitored host in my “~/var/pnp4nagios/perfdata/” folder and ~/var/check_mk/rrd/ is empty.
I noticed that e.g. for Filesystem_*.rrd I get the error only for these 2 (but on many hosts):
Filesystem___growth.rrd
Filesystem___trend.rrd
Filesystem_*.xml ( mentioned by @andreas-doehler ) files are present for the hosts mentioned in the log and also 4 more rrds per filesystem-service, checkmk does not complain about being present:
Then you forgot to create the RRD rules inside your configuration.
Please first create two rules with “Configuration of RRD databases of services/hosts” and then do the convert again.
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed. Contact an admin if you think this should be re-opened.