Hi,
I've managed to successfully create a cacti environment. Within this environment, I'm predominently using external scripts to query info from remote hosts. Although my scripts are fairly basic, the time it was taking to complete a 5 minute cycle wasn't long enough. I installed spine and was able to speed things up to about 240 seconds. I noticed via the logs that my data input method was being processed multiple times for the multiple graphs on a single host. I'd have thought the poller would intelligently determine if the data needed has already been received.
Let's say I have one DIM returning 7 values. Three data templates are created and use this DIM (i.e. 3 remote hosts). 4 graph templates are created per data template. When the poller runs, I notice that there are 12 queries for the same data. I'd have expected only 3, one for each remote host.
I must have configured the graphs incorrectly but I can't figure out how to create them such that only one query per remote server is necessary.
Any pointers would be appreciated.
Multiple queries for the same data
Moderators: Developers, Moderators
- gandalf
- Developer
- Posts: 22383
- Joined: Thu Dec 02, 2004 2:46 am
- Location: Muenster, Germany
- Contact:
This is a known drawback. It's on the todo list but yet no clue.
Here's how to workaround (already posted every other month, sigh)
Please create only the first graph "as usual". This will create the only set of rrd files and the only single poller entry required.
For all other graphs related to the same target/script, please create graphs from Graph Management, giving host and graph as input. Then select the already existing rrd file/data sources that shall be used by the graph.
Reinhard
Here's how to workaround (already posted every other month, sigh)
Please create only the first graph "as usual". This will create the only set of rrd files and the only single poller entry required.
For all other graphs related to the same target/script, please create graphs from Graph Management, giving host and graph as input. Then select the already existing rrd file/data sources that shall be used by the graph.
Reinhard
Who is online
Users browsing this forum: No registered users and 2 guests