To Continue with Paul's point. I believe that there are allot of performance gains to be had by running perf and looking at the output to see where our algorithmic/data structure choices are wrong and fixing those first before diving in and rewriting for multi-threaded code.
donald On Fri, May 29, 2015 at 9:41 AM, Paul Jakma <[email protected]> wrote: > On Fri, 29 May 2015, Michael H Lambert wrote: > > I just want to make sure I've been following this email chain correctly. >> It seems that there is at least some consensus for a single daemon for all >> VRFs. Does this preclude a thread per VRF within the daemon? >> > > generally, no. > > For existing Quagga libzebra daemons, probably that's going to be such a > long effort to achieve, that there will be significant risks as to whether > it can succeed. Multi-daemon would be much easier I suspect (it's not like > VRF instances need to share any data, so there's no memory/performance > trade-off issues that would generally give threads any benefit over > multi-process - unlike multi-processing bgpd). > > regards, > -- > Paul Jakma [email protected] @pjakma Key ID: 64A2FF6A > Fortune: > You can fool all the people all of the time if the advertising is right > and the budget is big enough. > -- Joseph E. Levine > > > _______________________________________________ > Quagga-dev mailing list > [email protected] > https://lists.quagga.net/mailman/listinfo/quagga-dev >
_______________________________________________ Quagga-dev mailing list [email protected] https://lists.quagga.net/mailman/listinfo/quagga-dev
