On Mon, Feb 01, 2016 at 04:47:56PM +0000, David Harton (dharton) wrote: > Hi folks, > > I didn't see any follow up to this response. > > My original concern was rte_eth_stats_get() moving away from a more > conventional based definition (note, I believe Matthew Hall made an > interesting suggestion to follow a MIB based definition elsewhere). > However, if modifying that API is not desired then I'd really like to have > some feedback about extending the current xstats model. > > Again, it is desired not to have to copy and/or parse strings for > scalability reasons but still maintain the "ABI flexibility" for which the > xstats model was geared towards. > > Thanks, > Dave
For me, I'd like to be able to get the core common stats in single memory blocks which are as self-consistent and atomic as possible. I'd prefer to only resort to stuff like xstats using some kind of string-to-number lookup at the beginning, and only for weird stuff not the common MIB-like items. >From past experience this would be very valuable. Matthew.