Help with debugging graph with holes.

[rvt]

Hi All,

I have a graph showing how many users are provisioned and for the life of me I can't figure out why it's only intermittently working.

I'm running Boost and I can see the update line in the boost.log:

Code: Select all

2012:02:25 13:17:13 - RRD Command 'update /opt/cacti-0.8.7i/rra/76/2775.rrd --template prov_subs  1330136042:13280315'
2012:02:25 13:17:13 - RRDUpdate OK Message: 'OK'
2012:02:25 13:17:13 - Host Disconnect Request Received. '127.0.0.1'

When I run the line manually using 'updatev' initially I got:

Code: Select all

[root@server1 log]# rrdtool updatev /opt/cacti-0.8.7i/rra/76/2775.rrd --template prov_subs  `date +%s`:13280315
return_value = 0
[1330136100]RRA[AVERAGE][1]DS[prov_subs] = NaN
[1330136700]RRA[AVERAGE][1]DS[prov_subs] = NaN
[1330136100]RRA[AVERAGE][1]DS[prov_subs] = NaN
[1330136700]RRA[AVERAGE][1]DS[prov_subs] = NaN
[1330136100]RRA[AVERAGE][1]DS[prov_subs] = NaN
[1330136700]RRA[AVERAGE][1]DS[prov_subs] = NaN
[1330136100]RRA[AVERAGE][3]DS[prov_subs] = NaN
[1330136100]RRA[MAX][1]DS[prov_subs] = NaN
[1330136700]RRA[MAX][1]DS[prov_subs] = NaN
[1330136100]RRA[MAX][1]DS[prov_subs] = NaN
[1330136700]RRA[MAX][1]DS[prov_subs] = NaN
[1330136100]RRA[MAX][1]DS[prov_subs] = NaN
[1330136700]RRA[MAX][1]DS[prov_subs] = NaN
[1330136100]RRA[MAX][3]DS[prov_subs] = NaN
[1330136100]RRA[LAST][1]DS[prov_subs] = NaN
[1330136700]RRA[LAST][1]DS[prov_subs] = NaN
[1330136100]RRA[LAST][1]DS[prov_subs] = NaN
[1330136700]RRA[LAST][1]DS[prov_subs] = NaN
[1330136100]RRA[LAST][1]DS[prov_subs] = NaN
[1330136700]RRA[LAST][1]DS[prov_subs] = NaN
[1330136100]RRA[LAST][3]DS[prov_subs] = NaN

Ran it again a few seconds later and it seemed to go through successfully:

Code: Select all

[root@server1 log]# rrdtool updatev /opt/cacti-0.8.7i/rra/76/2775.rrd --template prov_subs  `date +%s`:13280315
return_value = 0

Here's an example of 'rrdtool fetch <rrd> AVERAGE':

Code: Select all

1330125600: 1.3279340000e+07
1330125900: 1.3279331090e+07
1330126200: 1.3279318087e+07
1330126500: 1.3279344820e+07
1330126800: 1.3279362880e+07
1330127100: nan
1330127400: nan
1330127700: nan
1330128000: nan
1330128300: nan
1330128600: nan
1330128900: nan
1330129200: 1.3279522000e+07
1330129500: 1.3279533880e+07
1330129800: 1.3279534000e+07
1330130100: 1.3279596000e+07
1330130400: 1.3279635600e+07
1330130700: 1.3279636000e+07
1330131000: 1.3279677000e+07
1330131300: 1.3279715740e+07
1330131600: 1.3279716000e+07
1330131900: 1.3279778000e+07
1330132200: 1.3279778000e+07
1330132500: 1.3279778000e+07
1330132800: 1.3279844000e+07
1330133100: 1.3279881747e+07
1330133400: 1.3279928693e+07
1330133700: 1.3279958000e+07
1330134000: nan
1330134300: 1.3280042740e+07
1330134600: 1.3280080000e+07
1330134900: nan
1330135200: nan
1330135500: nan
1330135800: nan
1330136100: nan
1330136400: nan
1330136700: nan
1330137000: nan

The datasource has a max limit of 1000000000, is a gauge and the returned results always fall inside the valid range. Step is 300 and heartbeat is 600.

Any ideas? Thanks.