| Main Archive Page > Month Archives > syslog-ng-users archives |
Martin Holste <mcholste@gmail.com> writes:
> Yep, I think you're on the right track in that some rewriting will
> definitely be necessary for Mongo.
I'm hoping that it will be useful for JSON aswell - but that might need
some more work, to be really useful.
(The $(format-json) template function only supports a single level of
values, embedded objects aren't supported at all, yet, key rewriting
would be more useful if they'd be, like in Mongo's case.)
> I'm a bit concerned with performance, but Mongo will probably be the
> bottleneck when things don't fit in RAM anyway.
I'm not too worried about Mongo performance, there's many ways it can be
made to scale well.
The key rewriting will come with a cost of course (especially the
regexp-based one; I'm reasonably sure I can cook up prefix() to be quite
efficient), but I, for one, don't mind that all that much: my immediate
need is differently structured data, and I have enough free resources to
throw at the task. ;)
-- |8] ______________________________________________________________________________ Member info: https://lists.balabit.hu/mailman/listinfo/syslog-ng Documentation: http://www.balabit.com/support/documentation/?product=syslog-ng FAQ: http://www.campin.net/syslog-ng/faq.html