[ARGUS] argus collector
carter at qosient.com
Fri Sep 11 09:05:15 EDT 2020
Hey Riccardo !!
Yes, radium is the best way for real-time collection.
Configure your argus servers to listen on port 561 …. Or run radium on your remote argus servers to manage the port.
Argus will support a very small number of clients ... radium is designed to support a large number of concurrent clients.
On your central collector, configure radium to attach to your remote collectors … using the RADIUM_ARGUS_SERVER variable(s) in your /etc/radium.conf file. It should manage getting the connection and maintaining it (reconnecting if the connection goes away) and collecting data.
Radium will merge all the streams from all its server connections, and generate a single stream that it can offer to clients on the central collector through its port (default 561). Then, any number of clients can attach to this radium, and process the aggregate stream of argus records. This is the best way to get parallelism for processing argus records.
To create a repository on the central collector, run the program rastream.1 and have it split the aggregate stream into an archive. So assume you configured the radium on the central collector to offer its data on port 561, to create ‘canonical’ 5m argus archive you would run:
rastream -S localhost:561 -M time 5m -w /path/to/archive/$srcid/%Y/%m/%d/argus.%Y.%m.%d.%H.%M.%D
This should get you started … send email if you have any problems !!!
> On Sep 11, 2020, at 8:46 AM, Riccardo Veraldi <Riccardo.Veraldi at cnaf.infn.it> wrote:
> I have argus installed on a bunch of servers and it writes locally
> argus.out files.
> I would like to collect argus data from all the different servers to
> only one storage collector location, with all traffic bundled together,
> is that possible using radium
> or in any other way ?
> thank you
> argus mailing list
> argus at qosient.com
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the argus