as far as I know, collectd does not support config reload of running
instance (but check if it was not implemented in last few years).

But there is relatively simple workaround which we are using - we have
split our setup to more daemons, e.g.
- writer instance (writes data to disk)
- thresholding instance
- collecting instances

Instances communicate together over local udp socket, but they can also
communicate over lan (unicast or multicast) or even over the internet.

We can then e.g. restart collecting instance without having to wait to
flushing data to disk.

You can even start a new collecting instance before shutting down old
one and in this way you can make sure there are no 'holes' in data.

Best regards

Dne 24.10.2018 v 09:38 Oleg Rosowiecki napsal(a):
> Hello,
> Can I trigger a change in a plugin's behavior (regardless whether it's a
> read or write plugin) using some collectd API?
> For example, I'd like to change a flag that influences the behavior of a
> plugin or change a configuration item item for my plugin. Does collectd
> provide an API for that? If not, then maybe it is possible to tell
> collectd to reload a plugin dynamically or re-read its configuration?
> I'll appreciate any suggestions.
> Thanks,
> Oleg
> _______________________________________________
> collectd mailing list
> collectd@verplant.org
> https://mailman.verplant.org/listinfo/collectd

collectd mailing list

Reply via email to