perfsonar development work

[Loukik Kudarimoti <>]

Hi Roman, all,

Jeremy and Charaka from the EGEE project have some questions about the 
data retrieved from RRD MA. I have answered the first one but not sure 
about the one on time values.

Can anybody please help?

Thanks,
Loukik.


Jeremy Nowell wrote:
> Hi Loukik,
>
> I'm just running a few test queries against different PerfSONAR 
> measurement archives while Charaka is away.  I've come up with a 
> couple of issues, that seem common to different NM-WG v2(2) PerfSONAR 
> services.
>
> The first is the consolidation function,
> http://wiki.perfsonar.net/jra1-wiki/index.php/Metadata_Key_Request_with_filtering 
>
> states that supported values for consolidationFunction are AVERAGE, 
> MIN, MAX and LAST.  The test we do tries the first three, and it works 
> for AVERAGE and MAX, but fails for MIN, giving the result:
>
> <soapenv:Fault>
>   <faultcode>soapenv:Server.userException</faultcode>
>   <faultstring>org.perfsonar.service.commons.exceptions.SystemException:
>     [unknown]: RRDStorageManager.rrdFetch: Cannot retrieve data from
>     rrdjtool library: RRDFETCH failed: the RRD does not contain an RRA
>     matching the chosen CF
>   </faultstring>
>   <detail>
>     <ns1:hostname 
> xmlns:ns1="http://xml.apache.org/axis/";>srv4</ns1:hostname>
>   </detail>
> </soapenv:Fault>
>
> Any ideas?
>
> The other problem is with the time.  If the query contains a time 
> range, eg
> <parameter name="startTime">1160486108</parameter>
> <parameter name="endTime">1160489708</parameter>
>
> corresponding to the last hour, then I get results 30 minutes apart like:
>
> <nmwg:datum timeValue="1160487000" value="4.335193277741976E7"/>
> <nmwg:datum timeValue="1160488800" value="3.9467713667480946E7"/>
> <nmwg:datum timeValue="1160490600" value="NaN"/>
>
> As you can see the last value returned is from the future, and so 
> doesn't contain a valid result...  I would expect the results returned 
> to be from times within the specified time range, but maybe you could 
> clarify.
>
> Thanks for your help,
> Jeremy