Graph created but sometimes blank

[ryukirider]

i dont know why my cacti graph sometime will blank, I am using :
Cacti Version 0.8.7b
Cacti OS unix
SNMP Version net-snmp
RRDTool Version RRDTool 1.2.x

Can any brother have the same problem as mine?

Please tell me how to slove ! thank you very much

[acasto]

Umm... this could be anything. Is it happening on all graphs at the same time? If so, check you cacti.log and look for problems.


Adam

[ryukirider]

I've checked the log but cant find any abnormal. It sometime happen on all graph .

[gandalf]

Please try 2nd link of my sig and see especially the MAX data source issue
Reinhard

[ryukirider]

Dear Reinhard, can u give me the link please

[abinidi]

I too am getting these issues where the graph is broken. I have checked the logs and sometimes I get:

03/28/2009 06:35:02 AM - CMDPHP: Poller[0] ERROR: A DB Exec Failed!, Error:'1062', SQL:"insert into poller_output (local_data_id, rrd_name, time, output) values (105, 'traffic_in', '2009-03-28 06:35:02', '1711609424')'

But yet the next time it polls, it inserts the data correctly, no errors. So I really doubt it is Mysql or anything, as most of the time it is able to insert correctly. I would think it is a code issue with CACTI.

Here is an example of one of my graphs I am seeing.

Anyone have a suggestion?
System:
Centos5.2 x64
Mysql 5
Php5
0.8.7d with latest patches

[anton.prus]

i have this problem too
and i dont use spine

increasing threads or processes of poller didnt make any effect on graphs

please suggest how to solve it




________________________________
found one warning in log -
03/31/2009 07:30:00 AM - POLLER: Poller[0] WARNING: Poller Output Table not Empty. Issues Found: 1, Data Sources: (DS[1231])

[abinidi]

Based on the lack of responses on this and other posts regardling this major bug in Cacti, I don't believe it is fixable

[anton.prus]

ANY HELP??? WHO IS IN CHARGE HERE

[rootfive]

I am getting the same error recently. There are gaps in graph from time to time and here are some cacti log error output

04/03/2009 08:40:16 AM - CMDPHP: Poller[0] ERROR: A DB Exec Failed!, Error:'1062', SQL:"insert into poller_output (local_data_id, rrd_name, time, output) values (66, 'traffic_in', '2009-04-03 08:40:15', '3273489')'
04/03/2009 08:40:16 AM - CMDPHP: Poller[0] ERROR: A DB Exec Failed!, Error:'1062', SQL:"insert into poller_output (local_data_id, rrd_name, time, output) values (66, 'traffic_out', '2009-04-03 08:40:15', '3371139818')'
04/03/2009 08:40:16 AM - CMDPHP: Poller[0] ERROR: A DB Exec Failed!, Error:'1062', SQL:"insert into poller_output (local_data_id, rrd_name, time, output) values (64, 'traffic_in', '2009-04-03 08:40:15', '3722388339')'

I do not know how to solve too

[BSOD2600]

Based on the lack of responses on this and other posts regardling this major bug in Cacti, I don't believe it is fixable

Random gaps in graphs often is a sign of high network latency or response from a device/query and a lack of any data returned. Follow that debugging guide earlier posted.

I am getting the same error recently. There are gaps in graph from time to time and here are some cacti log error output...

Manually run one of those sql commands -- what is the error? Does the cacti technical support page show everything is good?

[chloetey]

Yea yea.. me also have the same problem here.
from the log, i have this warning. not sure will it affected the graph?

01/09/2010 11:09:00 AM - POLLER: Poller[0] WARNING: Poller Output Table not Empty. Issues Found: 836, Data Sources: traffic_in(DS[756]), traffic_out(DS[756]), traffic_in(DS[757]), traffic_out(DS[757]), traffic_in(DS[758]), traffic_out(DS[758]), traffic_in(DS[759]), traffic_out(DS[759]), traffic_in(DS[760]), traffic_out(DS[760]), traffic_in(DS[761]), traffic_out(DS[761]), traffic_in(DS[762]), traffic_out(DS[762]), traffic_in(DS[763]), traffic_out(DS[763]), traffic_in(DS[764]), traffic_out(DS[764]), traffic_in(DS[765]), traffic_out(DS[765]), traffic_in(DS[766]), Additional Issues Remain. Only showing first 20

[TheWitness]

Goto Console->Settings->Poller and disable Process leveling. If you are, in fact, using 0.8.7b, you have either two crontab's or your poller is overrunning itself, which typically happens when you don't have enough memory to cache all of your RRDfiles. This won't typically happen until you have a few k of rrdfiles.

TheWitness