Thanks Baszi! Right now I have created separate-bu-identical redis destinations: destination d_redis1 {stuff} destination d_redis2 {exactly the same stuff} ... destination d_redisN {still the same stuff} so I can filter the sources and have different log statements send to each destination. Is there a more compact way to do this? (I'd love to have a single d_redis that creates several threads based on load so I could lose the filters and separate log statements) Thoughts? Jim ---- Balazs Scheidler <bazsi77@gmail.com> wrote:

Yup. All of them use an independent thread. Even with threading disabled, as we are using a synchronous api to redis. On Apr 28, 2015 2:34 PM, "Jim Hendrick" <jrhendri@roadrunner.com> wrote:

...

Hi,

Could someone help clarify something for me?

I am running with threading enabled, and I have separate redis destinations (three currently) all pointing to the same server. My thought was that each destination would use a separate connection and thread. I then filter log sources and send (roughly) 1/3 to each destination.

My question is this. Does each destination get a separate thread? (reading the documentation it left room for doubt).

Also - more generally is this a reasonable way to get better throughput to a redis data store?

(note: both syslog-ng and redis are running locally, so the connections use 'localhost')

Thanks! Jim

______________________________________________________________________________ Member info: https://lists.balabit.hu/mailman/listinfo/syslog-ng Documentation: http://www.balabit.com/support/documentation/?product=syslog-ng FAQ: http://www.balabit.com/wiki/syslog-ng-faq