wu-sheng commented on issue #2044: [Proposal] Make Zipkin and Jaeger formats analysis product available URL: https://github.com/apache/incubator-skywalking/issues/2044#issuecomment-447207811 > Show us some sample Zipkin trace. How to design a no wait analysis system must be base on that. No, I just say wait for the whole trace. I think you misunderstand. Let's say right now, the whole trace is buffer and waiting, then do the reformat to SkyWalking native trace format, right? Because we want to reuse the process of trace parse codes. But, actually, we don't really need that based on the new core. We have scope dispatcher now. The only thing needs to wait is RPC related spans, we need the client and server span match to build the relation scopes, right? This part is still waiting, no change and no chance to improve. In this proposal, other local spans could be passed and saved. Only some RPC related trace id and span id need to buffer in local and do timeout/match (Span info saved in cache for sure, no matter what it is). Also, no need to wait for the whole trace ready, just a pair of RPC spans is ready, the analysis is good to go for this part.
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
