Gaps in Graphs... Please help :)

[Potter]

We are trying to run poller intervals at 1minute. It mostly works great, although sometimes through out the day we get random gaps in the graphs

Has anyone seen this before? Have we missed something?

Attached is what the graphs look like. Sometimes its worse then this and other times it doesn't happen for hours.

When we see gaps in the graph we have a look in the Cacti Log File and can see the following alerts:

06/10/2015 09:55:01 AM - POLLER: Poller[0] WARNING: Poller Output Table not Empty. Issues Found: 808, Data Sources: traffic_in(DS[1029]), traffic_out(DS[1029]), traffic_in(DS[1030]), traffic_out(DS[1030]), dropped_out(DS[1035]), traffic_in(DS[1039]), traffic_out(DS[1039]), traffic_in(DS[1040]), traffic_out(DS[1040]), traffic_in(DS[1156]), traffic_out(DS[1156]), traffic_in(DS[1157]), traffic_out(DS[1157]), traffic_in(DS[1158]), traffic_out(DS[1158]), traffic_in(DS[1159]), traffic_out(DS[1159]), traffic_in(DS[1160]), traffic_out(DS[1160]), traffic_in(DS[1161]), Additional Issues Remain. Only showing first 20

06/10/2015 09:55:01 AM - POLLER: Poller[0] WARNING: Scheduled Task is out of sync with the Poller Interval! The Poller Interval is '60' seconds, with a maximum of a '300' second Scheduled Task, but 120 seconds have passed since the last poll!


Our current set up is as follows:

Operating System: Windows Server 2008 R2
Webserver: IIS 7
Cacti:0.8.8c
MySQL: 5.5
PHP: 5.5
RRDTool: RRDTool 1.2.x
Net-SNMP: NET-SNMP version: 5.5
Plugin: Settings, Loginmod, Thold

Poller Type: cmd.php
Poller Interval: Every Minute
Cron Interval: Every Minute
Maximum Concurrent Poller Processes: 100

[BSOD2600]

Those warnings look like the poller is going over its 60 sec window, which is bad. look into making any custom scripts more efficient and/or optimizing Cacti (mysql db tuning, enabling Boost plugin, etc).