[Orca-users] Re: HELP: Can't get Orca to correctly read in non-orcallator related data

[Sean O'Neill]

At 11:17 AM 1/11/2003 -0600, Sean O'Neill wrote:

Please ignore my previous email.  I think I know what the problem is - 
working on it.

>Long email ... sorry.
>
>I've done this before for Weblogic SNMP data and I can't figure out why 
>this stuff isn't working this time.
>
>I'm pulling in java garbage collection info from some Weblogic logs.  I've 
>got the data laid out as follows:
>
>timestamp   MNGCCount  MNTotMinNGC   MNProcTimeLeft   MN%Lost
>1039537260        1  5.405599  54.59440  9.009331
>1039543380        1  4.680467  55.31953  7.800778
>1039574040        1  6.515461  53.48453  10.85910
><snip>
>1042209840        1  9.844697  50.15530  16.40782
>1042209900        1  9.074269  50.92573  15.12378
>1042209960        2  22.36275  37.63724  37.27126
>1042210020        1  9.309143  50.69085  15.51523
>1042210080        1  9.341319  50.65868  15.56886
>1042210140        2  22.22149  37.77850  37.03583
>
>The timestamp column isn't individual probes.  Each row represents a 
>accumulation of data for each specific minute.  But there are gaps in the 
>data.  There are periods where FGCs don't occur for long periods of 
>time.  At the end of my data, FGCs occur every minute.
>
>My gcallator.cfg file looks basically just like my orcallator.cfg except 
>I've modified it specifically for the above data columns.  Both .cfg files 
>are tuned for 60 second data intervals.  Sample of one section of 
>gcallator.cfg:
>
>plot {
>title                   %g Full GC Count
>source                  gcallator
>data                    MNGCCount
>legend                  Full GC Count / minute
>y_legend                Number of Full GCs
>data_min                0
>}
>
>Then I run orca as follows:
>
># /usr/local/orca/bin/orca -once -verbose /usr/local/orca/lib/gcallator.cfg
>Orca version 0.264 using RRDs version 1.000381 at Sat Jan 11 10:54:49 2003.
>/usr/local/orca/bin/orca: warning: cannot open state file 
>`/home/soneill/gc/orca/rrd/gcallator/orca.state' for reading: No such file 
>or directory
>Creating orca.gif.
>Creating rrdtool.gif.
>Finding files and setting up data structures at Sat Jan 11 10:54:49 2003.
>/usr/local/orca/bin/orca: making directory 
>`/home/soneill/gc/orca/rrd/gcallator/gcallator_certification-app01'.
>/usr/bin/ps: not found  <<< I'm running on this FreeBSD. Ignore this. >>>
>Current running time is 0:00 minutes.
>Loading new data from group gcallator for certification-app01.
>   Read  3363 data points from 
> `/home/soneill/gc/orca/gcallator/certification-app01/gcmndata-000'.
>Flushing new data and updating PNGs from certification-app01.
>   Creating RRD 
> `/home/soneill/gc/orca/rrd/gcallator/gcallator_certification-app01/gauge_MNGCCount.rrd'.
>   Creating RRD 
> `/home/soneill/gc/orca/rrd/gcallator/gcallator_certification-app01/gauge_MNTotMinNGC.rrd'.
>   Creating RRD 
> `/home/soneill/gc/orca/rrd/gcallator/gcallator_certification-app01/gauge_MN_pct_Lost.rrd'.
>Saving state into `/home/soneill/gc/orca/rrd/gcallator/orca.state'.
>Updating PNGs.
>Creating HTML files in `/misc/solaris-local/apache2/docroot/gc/'.
>/usr/bin/ps: not found
>Current running time is 0:08 minutes.
>Orca has completed.
>
>Its says it read over 3000 rows of data - kewl - but the graphs are 
>empty.  This is a rrdtool info of one of the rrd files:
>
>$ rrdinfo info gauge_MNGCCount.rrd
>ksh93: rrdinfo: not found
>$ rrdtool info gauge_MNGCCount.rrd
>filename = "gauge_MNGCCount.rrd"
>rrd_version = "0001"
>step = 60
>last_update = 1072920180
>ds[Orca19990222].type = "GAUGE"
>ds[Orca19990222].minimal_heartbeat = 120
>ds[Orca19990222].min = 0.0000000000e+00
>ds[Orca19990222].max = NaN
>ds[Orca19990222].last_ds = "UNKN"
>ds[Orca19990222].value = 0.0000000000e+00
>ds[Orca19990222].unknown_sec = 0
>rra[0].cf = "AVERAGE"
>rra[0].rows = 12000
>rra[0].pdp_per_row = 1
>rra[0].xff = 5.0000000000e-01
>rra[0].cdp_prep[0].value = NaN
>rra[0].cdp_prep[0].unknown_datapoints = 0
>rra[1].cf = "AVERAGE"
>rra[1].rows = 1488
>rra[1].pdp_per_row = 30
>rra[1].xff = 5.0000000000e-01
>rra[1].cdp_prep[0].value = NaN
>rra[1].cdp_prep[0].unknown_datapoints = 23
>rra[2].cf = "AVERAGE"
>rra[2].rows = 1200
>rra[2].pdp_per_row = 120
>rra[2].xff = 5.0000000000e-01
>rra[2].cdp_prep[0].value = NaN
>rra[2].cdp_prep[0].unknown_datapoints = 83
>rra[3].cf = "AVERAGE"
>rra[3].rows = 1098
>rra[3].pdp_per_row = 1440
>rra[3].xff = 5.0000000000e-01
>rra[3].cdp_prep[0].value = NaN
>rra[3].cdp_prep[0].unknown_datapoints = 83
>
>Why would RRD flag my data as "unknown" - I haven't beaten up the RRD site 
>yet - I know RTFM - hoping someone here has a quick answer.
>
>Without showing everything I have (making a longer email), does anyone 
>have any ideas why this is happening ?  Could the "data cap" situation I 
>described above be contributing - can't RRD handle gaps in data correctly 
>- I think it can ?
>
>--
>........................................................
>......... ..- -. .. -..- .-. ..- .-.. . ... ............
>.-- .. -. -... .-.. --- .-- ... -.. .-. --- --- .-.. ...
>
>Sean O'Neill

--
........................................................
......... ..- -. .. -..- .-. ..- .-.. . ... ............
.-- .. -. -... .-.. --- .-- ... -.. .-. --- --- .-.. ...

Sean O'Neill